Back to Home / #linode / 2011 / 06 / Prev Day | Next Day
#linode IRC Logs for 2011-06-28

---Logopened Tue Jun 28 00:00:33 2011
00:02-!-skegeek [] has quit [Read error: Connection reset by peer]
00:03-!-skegeek [] has joined #linode
00:04-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
00:04-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
00:06-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
00:06<newbie>bytemask: thanks. hmm..blank but I suppose it would be since I just upgraded from MySQL 5.0.75 to 5.1.37...guess no clues there about the "ERROR 1577" message I got during Ubuntu 9.04 to 9.1 upgrade
00:06-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
00:10-!-Kebn [] has quit [Quit: Computer has gone to sleep.]
00:11<+linbot>New news from forums: CPU Spike in General Discussion <>
00:11<Deadpan110>sudo mysql_upgrade -u root -h localhost -p –verbose –force
00:12<Deadpan110>and... there is no Ubuntu 9.1 ...but there is an ubuntu 9.10
00:13-!-Kebn [] has joined #linode
00:14-!-byronb [] has joined #linode
00:16<Deadpan110>*Ubuntu versions: Ubuntu [2 digits from YEAR].[2 digits from MONTH] ... eg: 10.10, 11.04, 11.10,12.04 etc
00:17<Deadpan110>so... April and October releases every year with a name that starts J J, K K, L L, M M, N N, O O, P P etc
00:17-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
00:17-!-Tigeda [] has quit [Quit: Tigeda]
00:18-!-Tigeda [] has joined #linode
00:18-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
00:20<Peng>Is it a bad idea to use different rDNS names for a box's IPv4 and IPv6 addresses?
00:20<newbie>deadpan110: thanks...the site seems to run ok anyway?
00:21<nohh>Peng: not sure why it'd ever matter?
00:21-!-xt3mp0r [~xt3mp0r@] has joined #linode
00:21<Deadpan110>Peng, I wouldn't think so - I tend to add rDNS for all my important IP's based on their job (even load balanced ones)
00:22<Peng>nohh: I dunno -- 's why I'm asking. :)
00:22-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
00:24<Kyhwana_>Peng: not realy
00:24<nohh>is there a reason you'd want them to be different though?
00:25<akerl>nohh: Because they're different IPs?
00:25<nohh>different IPs for the same machine/resource
00:26<Deadpan110>yep... i have mx1.mydom.tld - mx2.mydom.tld and an ip that is shared... mail.mydom.tld and then several IPv6 ...all different IP's to names
00:26<dcraig>I don't see how it's any different than having multiple ipv4 ips and having different rdns for those
00:27<dcraig>but if you're using ipv4 and ipv6 as two different ways to reach the same services, why would you want different rdns hostnames?
00:28-!-sivy [] has joined #linode
00:29-!-Kebn [] has quit [Quit: Computer has gone to sleep.]
00:30<newbie>deadpan110: just did "sudo mysql_upgrade -u -h localhost -p --verbose --force" - big spew of output...would it be right to assume that all the errors are summarised at the bottom of the output?
00:34-!-byronb [] has quit [Ping timeout: 480 seconds]
00:34-!-stafamus [~stafamus@] has quit [Ping timeout: 480 seconds]
00:40-!-sivy [] has quit [Remote host closed the connection]
00:47-!-byronb [] has joined #linode
00:47-!-xt3mp0r_ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
00:47-!-xt3mp0r [~xt3mp0r@] has quit [Read error: Connection reset by peer]
00:51-!-xt3mp0r [~xt3mp0r@] has joined #linode
00:52-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
01:03-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
01:04-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
01:04<Deadpan110>newbie, that has --verbose set, so it should spit out a lot of detail ... the command should fix your mySQL error caused by the upgrade from 5.0 upward
01:05-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
01:05-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
01:07<Peng>Whoops, I was AFK. Anyway, the reason I want different rDNS hostnames is so that I can pull the AAAA record from the main hostname without breaking FCrDNS.
01:08<bob2>what is FCrDNS
01:09<bob2>I'd make the 'default' ipv6 address's rdns match ipv4's
01:09<Peng>Really, the only time I ever did that was due to tunnel issues, or when transitioning from a tunnel to native, so I could probably stop doing it.
01:13-!-Hoggs [] has joined #linode
01:14<Peng>On a related subject, if you've taken a service or box down, and aren't using the IPs for anything else yet, do you leave the RDNS entries up?
01:14<Kyhwana_>might as well
01:14<bob2>I take them down
01:15<Kyhwana_>Or change them to ""
01:16<Peng>Some things only log rDNS hostnames, rather than IP addresses, so I feel guilty about breaking them... OTOH, things like that aren't very smart!
01:16<bob2>logging rdns that isn't fc'd is stupid
01:16<bob2>who does that
01:17<Peng>Right, that's *really* dumb. I meant forward-confirmed or not.
01:19-!-JSharp [] has quit [Ping timeout: 480 seconds]
01:22<bob2>Peng: have you managed to get RS to enable ipv6 for you?
01:23<newbie>If MySQL 5.1.37 complains after a 5.0.75 upgrade that a table is needs to be upgraded but still seems to work with that table in the part of the website that uses it - how's that happen? Is it really needing repair?
01:23-!-aussiekid [] has joined #linode
01:23<aussiekid>hi all
01:23<bob2>depends what it really said
01:23<bob2>maybe it said it upgraded it
01:23<navi>newbie: I'd be scared
01:23<navi>newbie: YOu might find really weird or quirky behaviour with a feature
01:24<Peng>bob2: I haven't tried yet. I only have a Cloud Server, which they haven't rolled out IPv6 for yet, and I'm not going to enable it on that immediately anyway.
01:24<bob2>support had no idea how to enable ipv6 for cloudfiles
01:24<newbie>bob2: it's an InnoDB table so the usual repair table won't cut it. It would seem that I needed to have done a mysqldump prior to MySQL upgrade then "mysql < dump.sql" after the upgrade. Missed that boat.
01:25<bob2>and despite a 17-CNAME chain, my cloudfiles hostname has no AAAA
01:25<Peng>I hope 17 is a joke?
01:25<newbie>navi: not so far but point taken..I don't like seeing anything that hints at error either
01:25<bob2>well, 4
01:25<bob2>but at least the old rfcs said more than one was trouble
01:25<Peng>bob2: Some user said that, for Load Balancers, support chat said it required a ticket.
01:25<bob2>Peng: ah
01:26<Deadpan110>newbie, the error is a well known one, Ubuntu did not upgrade correctly, hence you run the command with --force (just in case the db has been flagged as upgraded)
01:26<Peng>I think I have a chain of 4 CNAMEs, but it's for a case that will never be hit ever.
01:26<newbie>deadpan110: I did run with --force..that's how I noted the error in the first place.
01:27<Peng>bob2: 2 CNAMEs is very common now, thanks to Akamai, and I've seen them go up to 3. I would worry a bit about 4, though.
01:27<Peng>I'm sure DJB is rolling in his...bed?
01:27<Deadpan110>this link: ,,,seemed like the best candidate i found on google
01:27<bob2>oh, I lie, 3 CNAMEs then an A
01:28-!-JSharp [] has joined #linode
01:28<newbie>deadpan110: yeh I did that one..that's how I noticed the problem in the 1st place
01:29<newbie>if repair table doesn't work with an InnoDB table, are there any problems switching it to a MyISAM table, repairing it and then switching it back to InnoDB (apologies if that was a totally idiotic suggestion)
01:29<bob2>the 'switching' step is where you lose all your data
01:30-!-kevin [] has joined #linode
01:30-!-smsfail [] has quit [Remote host closed the connection]
01:31<newbie>bob2: thought that would be too simplistic. least for MySQL 5.0.75 to 5.1.37, no way around it for InnoDB tables needing repair but to mysqldump it, do the upgrade then hen "mysql < dump.sql" after the upgrade?
01:32<bob2>I don't know
01:32<bob2>like I said, I live in the myisam ghetto
01:32<newbie>bob2: maybe it isn't a ghetto after all given what I'm noticing here
01:33<bob2>myisam is terrible
01:34<@heckman>Myisam is by default, yes?
01:34<newbie>above mysqldump then mysql upload idea coming from
01:34<@heckman>newbie: There's a good instruction in the Linode Library...
01:34<@heckman>!library mysqldump
01:34<+linbot>heckman: 1. Back Up Your MySQL Databases -
01:34<newbie>heckman: I believe so but heard all sorts of goodness about InnoDB..hadn't come across needing to repair a table until now though
01:35<bob2>heckman: depends on the mysql version
01:35<@heckman>bob2: I think with Debuntu it use myisam by default
01:35<bob2>til 5.6
01:35<bob2>hopefully by then everyone uses drizzle or something isntead
01:37<StevenK>Or Postgres? :-)
01:38<newbie>heckman: I regularly use mysqldump via cron but what I wonder about now is that since "repair table" doesn't work on InnoDB tables..perhaps I should stick to MyISAM? Come to think of it...when you need to use repair table, does it preserve data entered after the
01:38<newbie>error which required it occurred?
01:39<bob2>depends how fucked the table is
01:39<bob2>afaik mysqldump doesn't work/works poorly for innodb anyway
01:42<newbie>deadpan110: thanks for link. any ideas about repairing just one table? my DB is running like there's no problem actually but there are 3 InnoDB tables that report they need to be upgraded
01:43<newbie>bob2: I must say it has worked ok so far for me but then again, I haven't had any disasters...your recommendation instead of mysqldump for InnoDB? Or since you're a MyISAM man, no opinion on that?
01:45<bob2>I gather percona's thing is the thing to use
01:45-!-JSharp [] has quit [Ping timeout: 480 seconds]
01:48-!-groggr [] has joined #linode
01:49<newbie>bob2: thanks..will look into percona.
01:49<newbie>groggr: hi
01:49<groggr>what's up?
01:49<groggr>i wish linode did hourly billing like rackspace
01:49<groggr>but that will never happen =[
01:49<bob2>I want a pony
01:49<bob2>I wouldn't say never
01:49<groggr>bob2: you think it's possible?
01:50<groggr>hrm, when might that happen?
01:50<newbie>so no 1 has an elegant solution for repairing just 3 InnoDB tables?
01:50<bob2>no idea
01:50<bob2>newbie: 'restore from backup'
01:50<bob2>I just think saying "that will never happen" is stupid
01:51<groggr>does anybody have at least 1 linode in each datacenter?
01:51<bob2>lots of people do
01:52<groggr>that's expensive!
01:52<bob2>compared to what?
01:52<kenyon>a box of cereal
01:52<bob2>oh, good point
01:52<bob2>it is also more expensive than one mintie per week
01:52-!-shirro [] has joined #linode
01:52<groggr>compared to godaddy hosting
01:52<groggr>it's expensive
01:52<bob2>yeah, but then you'd have godaddy hosting
01:52<groggr>i use godaddy hosting
01:52<bob2>so you'd have two problems
01:53-!-byronb [] has quit [Quit: Leaving...]
01:53<groggr>godaddy is one of the biggest internet companies
01:53<groggr>so they can be trusted
01:53<bob2>RS are pretty good for per-hour billing if that's what you're after
01:53<bob2>godaddy of course can't be trusted
01:53<groggr>why is that?
01:53<kenyon>plus they have danica patrick
01:53<groggr>SHE'S HOT!
01:53<bob2>their owner killed an elephant
01:53<groggr>i want to kill and elephant
01:53<bob2>that's healthy
01:54<newbie>bob2: way around it then..looks like I need to start from the beginning...might go all the way to Ubuntu 10.04 & doco. that then start again to make sure I've got it all accurately noted
01:54<bob2>also, godaddy vps hosting appears to be exactly the same price as linode
01:54<groggr>to want to kill and elephant is healthy? or actually killing and elephant?
01:54<bob2>oh, I lie
01:54<bob2>groggr: both
01:54<groggr>i watched the video lol he used a .50 rifle to shoot the elephant in the brain
01:55<groggr>elephants are supposed to be some of the most intelligent animals out there
01:55<bob2>this might be more on-topic in #batshit-teenagers
01:55<groggr>but that elephant didn't know to run when he had a gun aimed at his head
01:55<bob2>plus they can just put you on the watchlist immediately
01:55<groggr>which list?
01:55<groggr>the "internet ballers that enjoy killing elephants" list
01:56<groggr>i've been on there for years
01:56<groggr>right near the top
01:58-!-groggr [] has quit [Quit: leaving]
02:00-!-zcid [] has joined #linode
02:00-!-xt3mp0r [~xt3mp0r@] has joined #linode
02:00-!-Kyhwana [] has joined #linode
02:01-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
02:02-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
02:02<kenyon>I actually saw a load of low in manager.linode this morning. first time I've seen it not say idle.
02:03-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
02:04-!-groggr [] has joined #linode
02:04<groggr>which datacenter blocks 6667 again?
02:04<groggr>texas, right?
02:04<Peng>Don't insult my beloved Texas by saying such things! :(
02:05<groggr>i don't see the purpose of blocking 6667
02:05<groggr>anybody running an ircd maliciously without the machine owners knowledge would use a nonstandard port anyways
02:06-!-DJ_ [] has joined #linode
02:06<newbie>deadpan110: I note you wrote that "ubuntu didn't upgrade correctly" with regards to MySQL 5.0.75 to 5.1.37 when I upgraded from Ubuntu 9.04 to 9.10 - what exactly was that? The 3 tables which required "upgrade" check out ok in the origianl linode and I can't see what distiinguishes them to have a problem compared to all the others in the DBMS
02:06<groggr>lol you're still on ubuntu 9.04?
02:06<newbie>groggr: not for long
02:06<groggr>guys, come on. ubuntu is a desktop OS, stick to debian for servers.
02:06-!-Kyhwana_ [] has quit [Ping timeout: 480 seconds]
02:07*encode sticks to Windows for servers, and leaves linux to the desktop
02:07<groggr>i never understood the purpose of using ubuntu for a server OS
02:07-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
02:07<DJ_>ungh time out already
02:07-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
02:07<groggr>even if you use ubuntu on your desktop, debian is almost exactly the same
02:08<groggr>except a bit more stable
02:08<DJ_>+1 Debian
02:09<DJ_>8 minute timeout?
02:09<groggr>newbie: your nick is *extremely* fitting
02:09<DJ_>Christ almighty
02:09<newbie>groggr: that's why I picked it..and your's?
02:09<groggr>mine is completely random
02:10<newbie>enough said then :)
02:10<bob2>-> #stupid-trolling
02:10-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
02:10-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
02:10<+linbot>urmom is *always* relevant.
02:11<nohh>oh linbot, you troll
02:11<echoline>linbot is the ginormousest troll
02:11-!-djg320 [] has quit [Ping timeout: 480 seconds]
02:11<echoline>and lurker
02:11-!-DJ_ is now known as djg320
02:11-!-seanh-ansca [] has quit [Quit: Leaving.]
02:12-!-djg320-server [] has joined #linode
02:12<navi>Jr zhfg cergraq jr nyy hfr jvaqbjf freire ba bhe yvabqrf...
02:12-!-zcid [] has quit [Quit: CGI:IRC]
02:12<+linbot>echoline: (rot13 <text>) -- Rotates <text> 13 characters to the right in the alphabet. Rot13 is commonly used for text that simply needs to be hidden from inadvertent reading by roaming eyes, since it's easily reversible.
02:12<groggr>.voteban navi
02:12<nohh>quit trying to summon cthulu
02:12<Peng>djg320: If your registered your nick you could ghost it.
02:13<echoline>nohh: FINE
02:13<djg320>hehe, hadn't considered, been a long while since i've touched irc. thank you :)
02:13<navi>groggr: =(
02:13-!-groggr [] has quit [Quit: leaving]
02:14-!-groggr [] has joined #linode
02:14-!-djg320-server [] has quit []
02:15-!-djg320-server [] has joined #linode
02:15<djg320-server>need to get back in touch with irssi
02:16*echoline typitty type typos
02:17<echoline>djg320: stay on topic
02:17<echoline>djg320: tell us about your mom
02:17<@Praefectus>echoline: take your own advice
02:18<echoline>Praefectus: linode linodey linoder linode.
02:18*echoline afk to troll irl
02:20-!-atan3 [] has quit []
02:22<marius>What an anoyign individual
02:22<marius>Clearly a troubled child with supressed oedipusian tendencies towards his father.
02:22<echoline>marius: yeah, but he's an op. what can ya do
02:23-!-TIBS01 [~TIBS01@] has joined #linode
02:23*djg320 gasps
02:23-!-smsfail [] has joined #linode
02:23-!-smsfail [] has quit [Remote host closed the connection]
02:23<echoline>djg320: don't worry, i have a ginormous ego
02:24-!-echoline is now known as egoline
02:24<djg320>very nice
02:24<egoline>thank you
02:26-!-egoline is now known as echoline
02:27-!-djg320-server [] has quit [Quit: leaving]
02:27<djg320>the smart thing to do would be to use screen
02:27<echoline>is it smart?
02:27<echoline>screen sometimes seems to be a security risk
02:28<groggr>screen a security risk?
02:29<echoline>uriel said so
02:29-!-TIBS01 [~TIBS01@] has quit []
02:29<groggr>what was the reasoning uriel gave?
02:29-!-TIBS02 [~TIBS01@] has joined #linode
02:30<echoline>groggr: a swift kick to the nuts
02:30<echoline>seriously though...
02:30<marius>for it ot be a risk someone would have to get access to your box
02:30<echoline>it does seem a little insecure
02:30<marius>in which case your screwed regardless.
02:30<echoline>oh, my screwed?
02:31<echoline>dtach at least makes a file in whatever directory you're in
02:31*groggr sighgs
02:31*groggr sighs
02:31*groggr lol
02:31*groggr rofl
02:31<echoline>but i dunno
02:31<groggr>ah, i guess /me is ignored by mikegrb
02:31<echoline>i will have to ask uriel sometime
02:31*navi cake
02:32<navi>It's a new product, available on store shelves now
02:32<navi>A dessert with all the taste and nutritional value of navi.
02:32<echoline>mmm navi
02:32<@mikegrb>mmm cake
02:33<navi>Grr, this is really annoying me now
02:33*groggr lol rofl cake
02:33-!-djg320-server [] has joined #linode
02:33<newbie>mikegrb: seems to be your standard line :)
02:33<navi>I'm trying to draw a regression line through some data, just can't get one that matches the data
02:34<navi>Ah. I think I have an idea now
02:34<echoline>navi: i took a class which covered that and don't remember a damned thing
02:35<echoline>at least about that
02:35<navi>echoline: I'm going to do a crappy job at it but make it work nonetheless
02:35-!-Parallax [] has joined #linode
02:35<navi>To... my TI-83!
02:36*navi will just try all the points and try every regression line on it, and then see what R^2 he gets
02:36<echoline>you need at least 2 i thought
02:36<echoline>how can you have a line with one point?
02:37<navi>echoline: What?
02:37*echoline is stupid though
02:37<navi>echoline: No, I mean I'll try transposing all the points into my calculator
02:38<navi>echoline: As in, I won't just copy a representative set of the data, I'll put in every single point
02:38<echoline>that should work fine if you adjust the harmonic resonance of the subspace interpolation.
02:38<echoline>(since you mentioned data)
02:38-!-groggr [] has quit [Quit: leaving]
02:38<echoline>haha ^
02:38<echoline>what a grouch
02:39<echoline>!rot13 bibble
02:39<+linbot>echoline: ovooyr
02:39<navi>echoline: Why?
02:40<navi>Gnedhva Sva-gvz-yva-ova-juva-ovz-yvz-ohf-fgbc-S'gnat-S'gnat-Byr-Ovfphvgoneery
02:40<echoline>navi: what!? that meant something!?
02:40<Parallax>I think navi is trying to call forth Cthulhu
02:40<echoline>god dammit
02:40<navi>The first one was just a sound effect
02:40<navi>The second one is encoded, yes.
02:41<Parallax>is that Lovecraft encryption?
02:41<navi>It's a slow decaying unlucky number
02:41<echoline>am i going to be unlucky
02:42<djg320>!rot13 I am so confused
02:42<+linbot>djg320: V nz fb pbashfrq
02:42<djg320>what am I doing up
02:43<navi>My encoded text was from Monty Pythin
02:43<navi>Lovingly known as
02:44<navi>and reads "Tarquin Fin-tim-lin-bin-whin-bim-lim-bus-stop-F'tang-F'tang-Olé-Biscuitbarrel"
02:44-!-Pupeno_ [] has joined #linode
02:44<navi>Although I do prefer the other name
02:45<navi>Malcolm Peter Brian Telescope Adrian Umbrella Stand Jasper Wednesday (pops mouth twice) Stoatgobbler John Raw Vegetable (whinnying) Arthur Norman Michael (blows squeaker) Featherstone Smith (whistle) Northgot Edwards Harris (fires pistol, then 'whoop') Mason (chuff-chuff-chuff-chuff) Frampton Jones Fruitbat (laughs) (squeaker) Gilbert (sings) 'We'll keep a welcome in the' (three shots) Williams If I Could Walk That Way Jenkin (squeaker) Tiger-drawers Pra
02:45-!-mcinerney [] has quit []
02:45<navi>Now /that's/ a name.
02:45<skegeek>This may be dumb, but would it be feasible to create a 'mini vps' service on a linode server offering sort of like a subset of what linode offers?
02:46<Parallax>Xen inside Xen?
02:46<navi>skegeek: You can't run Xen on Xen, but from what I hear you might be able to run OpenVZ
02:46<djg320>the overhead
02:47<navi>skegeek: The extra layer might cause performance issues, and you'd not be able to split a 512 into two 256s because of the kernel on the box taking some up, you'd end up with like... 2 200s...
02:47<Deadpan110>just resell linodes and manage them for the customer!
02:48<navi>And it wouldn't really be cost effective from what I can tell, if someone wanted a 200, they wouldn't pay $10 for it really
02:48<skegeek>I was thinking of the "virtualization" created using the same softwares a typical vps would have, such as servers for weband ssh etc.
02:49<Peng>skegeek: I think it's a bad idea. Virtualization inside virtualization kind of sucks.
02:49<navi>You could probably make better money being a shared web host or something
02:49<djg320>overhead + disk i/o = brutal
02:50<navi>Give ten people 1GB of disk and 10GB bandwidth for $3/mo and you're laughing.
02:50<Peng>!dns6 2600:3c00::2:b001
02:50<Peng>I stopped using different rDNS entries. :-\
02:50<skegeek>I wasn't thinking of actually running virtualization inside virtualization, just mimicking the virtualization with smaller offerings.
02:50<navi>You could sell it as "We run on trusted Linode servers and we promise to always keep software up to date"
02:51<navi>That's one thing I hate about shared hosting
02:51<Peng>skegeek: How do you "mimic" virtualization?
02:51<navi>"Oh, we only use PHP 4 because some of our customers don't want to upgrade..."
02:51<djg320>skegeek: You could offer shell access?
02:52<djg320>though it wouldn't be very logical without several IPs
02:52<skegeek>navi: PHP4 and 5 can be run side-by-side
02:52<navi>Would linode accept that as a reason?
02:52<skegeek>I doubt it.
02:52<navi>A request for a second IP "Because I want to virtualize and put an IP on each instance"
02:53<djg320>Case and point - justification is a bitch.
02:53-!-advion- [] has joined #linode
02:53-!-advion [] has quit [Read error: Connection reset by peer]
02:53*djg320 clings onto ipv4
02:53<skegeek>It's a poor excuse, don't cut future business because some current customers will leave. Take the bigger customer base and tell the whiners to find another way home.
02:55<djg320>Those afraid to innovate and adapt will die.
02:55<Peng>skegeek: Wait, who's excusing what?
02:55<skegeek>I upgraded to php5 as soon as I knew it was out.
02:58-!-TIBS02 [~TIBS01@] has quit [Ping timeout: 480 seconds]
02:59-!-skegeek [] has quit [Quit: HydraIRC -> <- \o/]
02:59-!-djg320-server [] has quit [Read error: Connection reset by peer]
02:59<djg320>Generally not a good sign
02:59<dcraig>what'd you do, djg320?
02:59-!-xt3mp0r [~xt3mp0r@] has joined #linode
03:00<djg320>hosed something
03:00-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
03:00<djg320>made a mistake in /etc/network/interfaces
03:00<dcraig>better fire up lish
03:01<+linbot>LISH allows you to perform certain actions without having to log in to the Linode Manager. LISH's primary function is to allow you to access your Linode's console, even if networking is disabled.
03:01<djg320>All over it, <3 lish
03:01*dcraig slaps heckman around a bit with a large southern sandfish
03:01*heckman slaps dcraig around a bit with a large cactus
03:01-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
03:01-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
03:02-!-djg320-server [] has joined #linode
03:02*djg320-server revives
03:03<djg320-server>newark230 = the bomb
03:03<dcraig>man my newark linode doesn't even have a quad core :p
03:03<Zr40>dcraig: afaik all of them have quad cores
03:03<djg320>Really old 'node?
03:04<dcraig>maybe they compensate by only putting 20 users on this box though :)
03:04-!-jaidan [] has joined #linode
03:04<dcraig>Zr40, now you know differently!
03:04<dcraig>I kinda like it
03:04<dcraig>it's special
03:05<djg320>have a Linode 1536 and 512
03:05<Peng>dcraig: What CPU does it have?
03:05<djg320>omg my home storage server just took off
03:05<djg320>cpu fan is pegged, wtf is it doing
03:05<Peng>Oh wow. That's ooold!
03:05<djg320>be back momentarily
03:05<dcraig>storing things, obviously...
03:05<Peng>dcraig: You're probably the only person left on the box.
03:06-!-Big-Papa [] has quit [Quit: Changing server]
03:06<djg320>ungh windows update
03:06<djg320>...go figure.
03:06*Praefectus moves 80 linodes to the host dcraig is on
03:06<djg320>server 2008 r2, opinions?
03:06<Peng>Man, a 5130.
03:07<dcraig>I bet Praefectus won't give us the real stats on how many users are on newark247
03:07<Peng>I'm pretty sure I was on something better 2 years ago.
03:07<bob2>over 9000
03:07<@Praefectus>its a 512 host, the stats on how many users is listed on the main site
03:07<Peng>Praefectus: That wasn't the question.
03:08<@Praefectus>well 40+80 = 120
03:08<dcraig>site lists averages
03:08-!-djg320-server [] has quit [Quit: leaving]
03:08<@Praefectus>its gonna take me a while to move those 80 cuz ima look for the cpu and io hogs
03:08<dcraig>we already know a "typical" 512 host has about 40
03:08<djg320>model name : Intel(R) Xeon(R) CPU L5420 @ 2.50GHz
03:09-!-jaidan [] has quit [Quit: CGI:IRC (Ping timeout)]
03:09<djg320>I want one >.>
03:09<Zr40>Praefectus: hmm, where's that listed?
03:09<@heckman>Sounds like an AOL chat room in here...
03:10<bob2>how many slices is that?????
03:10*heckman slaps bob2 around a bit with a large cactus
03:11<Peng>Praefectus: The question was how many nodes are on that specific host. Given the age of the CPU model, it's a reasonable assumption that it's an old host, and it's a reasonable assumption that it's being phased out and has fewer than the typical number of nodes on it. Refusing to answer is fine; but don't evade the question. :P
03:11<djg320>i want to bump up to a 2048
03:11<Zr40>so... a 20 GB linode would have the server all for itself?
03:11<Peng>Zr40: Probably.
03:11<bob2>Zr40: apparently not
03:11<Peng>It's unknown if Linode has hosts with 40+ GB of RAM.
03:12<dcraig>Praefectus has discovered it actually has the full complement of 40 nodes and doesn't want to confirm that :p
03:12<Peng>They most likely didn't at the time the 20 GB plan came out, but that may have changed since.
03:12<@heckman>I can give you a hint. They have over 9000 kB. :p
03:12<@Praefectus>dcraig: negative
03:12<bob2>48GB isn't very big for new servers
03:12<@Praefectus>but even if it did have all 40 on there, you would still have your listed resources
03:12<navi>heckman: The last time I was measuring things like that was when I was trying to decide what would fit on a floppy disk
03:12<bob2>if you're already getting quad cpu quad core, it might even be the sweet spot for ram/$
03:12<dcraig>I'm not complaining, it works great! :D
03:13-!-kedarm [] has quit [Ping timeout: 480 seconds]
03:13<navi>How times have changed
03:13<navi>Now I look at a file and if it's under about 50MB, I really don't care how big it is
03:14<navi>(e.g. applications that take up under 50MB, I'll keep 'just in case I want to use it')
03:15<Peng>dcraig: How old's the node?
03:15<Peng>dcraig: Or more to the point, how long have you been using that host?
03:15<dcraig>man I gotta find my invoices
03:15<djg320>top - 03:15:42 up 16 min, 1 user, load average: 0.03, 0.03, 0.01
03:15<djg320>you dig my uptime, yes.
03:16<dcraig>november 23, 2010
03:16<dcraig>not that old
03:16<Zr40>nobody can beat 16 minutes!
03:16<encode>03:17:12 up -12 min, 0 user, load average: -5.00, -8.00, -1.12
03:17<encode>i can :)
03:17<Peng>Yeah, when I was on dallas62, it had 5130s. I moved off it in 2009-04.
03:17<Peng>It was UML.
03:17<Zr40>my atlanta linode has 5520s
03:17<navi>Do you think Linode would be happy if you were on one of those hosts and you asked if you could be moved onto a different host in the same facility with a better CPU?
03:17<Peng>I moved to, I believe, dallas153, which was on 5420s.
03:17-!-ktabic [] has joined #linode
03:18<djg320>encode: I'm intrigued
03:18<encode>djg320: why?
03:18<Peng>navi: Shrug. It would let them phase out the hardware sooner, but until they finally shut it down they'd be using *more* resources, since they have to keep the box on until the last user moves off.
03:18<encode>djg320: it's going down in 12 minnutes
03:18<navi>Peng: That was what I was thinking
03:18<encode>and when it comes back up, that will be the host load
03:18<dcraig>sometimes they force migrations to new hardware
03:18<encode>or something like that. interpret it how you will
03:19-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
03:19-!-newbie [] has quit [Quit: CGI:IRC]
03:19<Zr40>dcraig: does that involve any downtime?
03:19<Peng>dcraig: Moving to a new host involves downtime.
03:19<dcraig>it did, but I think it's very rare
03:19-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
03:19<Peng>D: ! My logs say we were talking about dcraig's 5130s in December!
03:19<navi>What is the best processor Linode uses in boxen?
03:20<encode>the A5
03:20<@Praefectus>pentium II's
03:20<navi>I mean, modernwise
03:20<dcraig>crap, we're doing repeats
03:20<navi>I assume by basic arithmetic of the ones mentioned so far, the 5520 (cuz its bigger)
03:20-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
03:20<dcraig>no more 5130 talk...just review your old logs :p
03:20<Peng>Last time I said "Woah, 5130? THat's ooold.". This time I said "Oh wow. That's ooold!".
03:21-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
03:21<Zr40>djg320: that's like, 2950 better!
03:22-!-Edgeman [] has quit [Remote host closed the connection]
03:23<Peng>navi: I think L5520s are the most recent ones I've heard of...
03:23-!-jaidan [] has joined #linode
03:23<navi>yay, that's what I've got in london
03:24<djg320>I want one in my house
03:24<navi>I already have the cloud in my house, I have no room for anything else
03:24<rnowak>puny little E5520
03:25<@heckman>Same. Sadly, the cloud makes the humidity pretty much unbearable.
03:25<djg320>not the chipset itself
03:25<djg320>a nice beefy 1U rack to do some...projects...with
03:25<rnowak>can't wait for ivy-bridge generation xeons to upgrade my workstation, om nom
03:26<djg320>i have quite possibly the two worst machines in one room for heat generation
03:27<djg320>small room, both running with any load = 81 F in a matter of minutes
03:27<rnowak>do supermicro's motherboards allow for overclocking of xeon CPUs?
03:27<Peng>overclocking D:
03:27<djg320>p4 prescott 3.06, amd x4 quad :|
03:27<Zr40>81 F?! that's like, cold
03:28<djg320>when the ambient temp is 70?
03:28<Zr40>yes. Tiny difference in temperature
03:28<rnowak>Peng: my two X5650 love their OC ;)
03:29<djg320>Well, I should be used to pretty much any temperature difference.
03:29<djg320>I live in Maine :|
03:29<Zr40>now, if it was 81 C, that's HOT
03:29<djg320>i don't mean chipset temp lol
03:30<djg320>i mean ambient room temp
03:30<Zr40>ah, I see
03:30-!-xt3mp0r [~xt3mp0r@] has joined #linode
03:30-!-advion- [] has quit [Quit: Leaving]
03:31-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
03:31<Peng>Yes, 81C room temperature would be a problem...
03:31-!-xt3mp0r_1 [~xt3mp0r@] has quit [Read error: Connection reset by peer]
03:31-!-xt3mp0r__ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
03:32<Peng>Attack of the xt3mp0r clones
03:32<djg320>p4 prescott, amd x4 quad, c2d laptop converted to debian server, personal use laptop is core i3, one router, 2 access points, couple switches
03:32<djg320>packed into a man cave
03:32<djg320>if you can call it that
03:32<rnowak>I have a p4 prescott as well which I don't use anymore, which could double as a heat radiator
03:32<djg320>very much so!
03:33-!-jaidan [] has quit [Quit: CGI:IRC (Ping timeout)]
03:33-!-jaidan [] has joined #linode
03:34<djg320>should be slapping the a/c in soon
03:34<djg320>but predictable weather around here is a joke
03:36<rnowak>I'm keeping RH at 45% and temperature at 19C, comfortable <3
03:37<Peng>19C!? Cold!
03:37<@heckman>I hope that's CPU temps.
03:38<rnowak>that's room temperature, the rack enclosure is well-ventilated as well though
03:38-!-jaidan [] has quit [Quit: CGI:IRC (Ping timeout)]
03:38<djg320>I could live with that
03:39<rnowak>cryolab at work is at 2C, and later this year we're getting a sub-zero lab <3
03:39<rnowak>(just found out today)
03:39<djg320>You hit 40 degrees in Maine and everyone runs out ina bikini
03:39<djg320>women preferably
03:42<tonyyarusso>40...C? That's hiding in the basement with a fan running weather.
03:42<djg320>40 F
03:42<tonyyarusso>Oh, okay :P
03:43<tonyyarusso>Then by all means break out the bikinis - the ice is gone.
03:43<tonyyarusso>No need to bring the ax to swim.
03:44-!-djg320-server [] has joined #linode
03:44<djg320>recommended apache2 alternative?
03:45-!-Jippi [~jippignu@] has joined #linode
03:45<Peng>Lighttpd? Nginx?
03:46<djg320>LEMP on Deb looks nice
03:46-!-kedarm [] has joined #linode
03:47<djg320>Linode Library = Amazing
03:49<MarkJ>Linode * = Amazing
03:51-!-Hoggs [] has quit [Quit: Leaving]
03:52<+linbot>urmom is amazing too!
03:53<Zr40>urmom is so amazing, linbot took three minutes to recover
03:53*djg320 coughs
03:55-!-djg320-ios [] has joined #linode
03:55<+linbot>New news from forums: network external storage in Sales Questions and Answers <>
03:56-!-SpaceHob1 [] has joined #linode
03:56<djg320-ios>Colloquy is nice
03:56<+linbot>Could also use SSH + tmux + irssi like a real winner!
03:58<Peng>linbot: As if you'd know!
03:58<djg320-server>prompt + irssi works well also
03:58<rnowak>"irssi connectbot"4lifeOnTheGo#1
03:59-!-spkitty [] has quit [Quit: spkitty]
03:59<+linbot>Actually I was just reiterating something I read in a previous conversation.
03:59<@heckman>Also, trying to TAB "mohawk" = failure. :<
03:59<Peng>linbot: You should never listen to what people people say on IRC. They're all idiots.
04:00<+linbot>Affirmative. Which is why I am not listening to you right now.
04:00<Peng>linbot: Excellent
04:00<rnowak>heckman: that patched connectbot makes two of the useless buttons on my physical keyboard on the phone into alt and ctrl, pure love <3
04:03<@heckman>I use Swype, so unfortunately I need to click.
04:04<@heckman>I've also not generated a pubkey for my phone....because effort.
04:04<mdcollins_>I love you guys and your random conversations. tmux looks cool.
04:04<rnowak>I use the same on my phone and laptop because effort
04:04-!-mdcollins_ is now known as mdcollins
04:05<rnowak>laptop's fs is encrypted, my phone's isn't though... :/
04:05<@heckman>rnowak: Well that makes sense. I'm not too worried about someone obtaining my desktop/netbook key. Good luck obtaining the passphrase.
04:05<rnowak>I am finding android a bit... "insecure2 in that regard, just like your email accounts and everything being directly accessible as soon as it boots up :|
04:05<@heckman>Well, you don't have to do that.
04:06<rnowak>not like there's a choice if you want to have gmail on the phone?
04:06-!-djg320-ios [] has quit [Quit: Colloquy for iPod touch -]
04:06<rnowak>I wish they'd get stored in the credentials store which can have a passphrase to unlock :/
04:06<dcraig>can you have a top secret 4-number pin?
04:06<@heckman>mdcollins: started using tmux about two months ago.
04:06<@heckman>dcraig: You can have a passphrase now.
04:06<rnowak>to unlock the phone, yes - but anyone with half a brain will be able to get past that
04:06<@heckman>Well, at least they have that on CM7...
04:06<Zr40>rnowak: simple workaround. phone missing? -> change password
04:07<dcraig>if linode ever gets hacked, I feel it'll be because lulzsec figured out heckman's pass is 'boobs'
04:07<@heckman>pfft, I'm more secure than that.
04:07<@heckman>It's b00b5...
04:07<dcraig>you just told everyone
04:07<Zr40>I wouldn't call using someone's password 'hacking'
04:07<SpaceHob1>dcraig: your vpn is your weapon. Without your weapon, you are 8008132
04:08<@heckman>dcraig: I thought that IRC automatically asterisked out your password?
04:08<SpaceHob1>Zr40: I wouldn't call DOS attacks "hacking" either
04:08<dcraig>well I saw the letters
04:08<rnowak>that was your password you saw, dcraig
04:08<Zr40>heckman: it's all asterisk for me. Perhaps your password is the same as dcraig's?
04:09<@heckman>hunter1 -- that is all
04:09<@heckman>Ugh, I need to stay awake for five more hours. -_-;;
04:09<Zr40>heckman: that's what #linode is for :)
04:10<@heckman>Can't type too loud. Girlfriend is attempting to sleep. SWEET! A/C kicked on and is drowning out the keyboard.
04:10<Zr40>heckman: local time?
04:10<dcraig>do you have a particularly loud keyboard?
04:10<rnowak>heckman: woerk tiem atm?
04:11<@heckman>rnowak: negative, it's my weekend. Work tomorrow night, however. Going to be a super-long evening as I need to drive back to where I'm from to get my car inspected.
04:11<Parallax>Why can't you go to sleep?
04:11<rnowak>heckman: oh
04:11<@heckman>So work, then a few hours of driving, and then yeah... dcraig: Nope, MacBook keyboard.
04:12<Zr40>heckman: those aren't particularly loud
04:12<@heckman>Parallax: I work from 10p to 9a tomorrow night. I need to sleep tomorrow.
04:12<rnowak>Y U NO SLEEP NAO
04:12<dcraig>wow caker's a slave driver
04:12<@heckman>Zr40: Well, no. But if you're two feet away they make some sound.
04:12<dcraig>11 hour shift?
04:12<@heckman>dcraig: four days a week. Best thing ever!
04:12<dcraig>do you get a bathroom break?
04:13<rnowak>he gets a bucket
04:13<@heckman>Yeah, was either I get a lunch break or need to use a bucket.
04:13*heckman chose bucket
04:13<dcraig>y'all should form a union
04:13-!-nisstyre [~nisstyre@] has quit [Ping timeout: 480 seconds]
04:13<dcraig>speak truth to power
04:13<@heckman>But seriously. I kinda was like "Damn, 11 hour days" when I first saw it. But after the first week I loved it.
04:14<dcraig>yeah it's not like you really "work" for those 11 hours
04:14<@heckman>Definitely not. When you love what you do it's not a job. :>
04:15<MarkJ>PHP is definitely a job then :p
04:15<Zr40>MarkJ: I'd call it torture
04:15<MarkJ>hehe, that's one way to look at it
04:16<dcraig>it's a poor workman who blames his tools
04:16<@heckman>I have some language relearning and some language learning ahead of me.
04:16<Peng>It's a poor workman who has crappy tools!
04:16<Zr40>dcraig: good workmen don't have tools working against him
04:17<MarkJ>I wanted to move to another platform for this project... and I would if I had to time to (re)learn anything else...
04:17<rnowak>good workmen use even bad tools to their advantage
04:17<MarkJ>yeah that's me rnowak . it doesn't make them suck less though!
04:17<Zr40>rnowak: only if there are no good tools available
04:17<rnowak>sometimes bad tools is what you got, and you can still make great things with them
04:18<MarkJ>I want to tinker with CF again
04:18<@heckman>CF definitely looks interesting.
04:18<Zr40>doesn't use CF?
04:18<rnowak>it does
04:18<@heckman>Our Dom0 kernels run entirely in CF. :>
04:18<Zr40>it's CF all the way down
04:18<rnowak>I thought it was EmacsOS heckman?
04:19-!-Eliteoomph [] has joined #linode
04:19<@heckman>We're actually intercepting your information exchange with your Linode and are converting your PHP to CF.
04:19<MarkJ>and major telco forum here in Aus is written in CF too ([forums.]
04:19<Eliteoomph>Hey heckman,... I have a tech question that isn't exactly linode related but it should be a pretty easy one and I was just wonering if you could help me out
04:20<@heckman>Probably would be best to ask in the channel...
04:20<MarkJ>hehe hecman
04:20<Eliteoomph>yeah thats what i mean ;o)
04:20-!-stijlfigurant [] has joined #linode
04:20<MarkJ>so just ask Elite
04:21<stijlfigurant>i got a question
04:21<stijlfigurant>i built this website for a client which is running on linode
04:21<stijlfigurant>now the client has a couple of domains
04:21<Eliteoomph>I have a domain registered at godaddy and i am trying to point the nameservers to my hosting account at justhost..... I have done this a dozen times with no problems but on this domain it is telling me my nameservers are not registered...? What does that mean>
04:21<stijlfigurant>pointing to the linode server
04:22<stijlfigurant>the problem is all those domains, collected in the past are at a couple of different servers
04:22<stijlfigurant>so the question is
04:22<MarkJ>What are you setting for the nameservers? ->
04:22<stijlfigurant>can all domains be put under linode instead?
04:22<stijlfigurant>so there would be one subscription and not 5
04:22<rnowak>stijlfigurant: uh?
04:22<dcraig>you can point as many domains at your linode IP as you'd like
04:22<marius>stijlfigurant: Linode isnt' a domai nregistrar, so no.
04:22<Eliteoomph>this is apart from my linode account
04:23<rnowak>Linode is not a registrar, but you can use their DNS service for any number of domains as long as you have an active linode.
04:23<dcraig>two domain questions at once is confusing :p
04:23<@heckman>Eliteoomph: Are you trying to run your *own* name servers?
04:23-!-djg320 [] has quit [Ping timeout: 480 seconds]
04:23<Eliteoomph>I am not trying I have them running on about 6 or 7 accounts
04:23<stijlfigurant>ok, so you'd suggest i get all the domains at one registrar instead
04:23<Eliteoomph>I never haver this problem that is why it is so strangr
04:23<marius>stijlfigurant: Indeed.
04:24<@heckman>Eliteoomph: Are you trying to run your own name servers.
04:24<stijlfigurant>any good options for such a registrar
04:24<marius>stijlfigurant: Everyone has different opinions, but they tend to bend towards and it seems
04:24<rnowak>Eliteoomph: you really should not run your own nameservers unless you know what you're doing.
04:24<dcraig>I use and
04:25<marius> is new to me *has a look*
04:25<@heckman>Eliteoomph: You need to have them recognized. Might wanna ask GoDaddy how to do it. I had to do something with my old registrar to get my registered. I forget what it was.
04:25<MarkJ>and living in Aus I use MDWebHosting (now part of NetRegistry) and VentraIP for my domain registrations
04:25<@heckman>I think there may have been a "Register Name Servers" URL at my old registrar.
04:25<Eliteoomph>.... i am confused I have half a dozen other domain on godaddy set up with these exact same nameservers
04:26<@ericoc>register nameservers link at godaddy, it was like hidden in a corner back when i used them
04:26<dcraig>Eliteoomph, what are the domains and nameservers?
04:26<@heckman>So *ask* GoDaddy?
04:26<@heckman>If they are giving you the error why are you not asking them?
04:26<marius>Does the domains have zones on yoru NS'es already ?
04:26<Peng>GoDaddy has good customer service, as I recall?
04:26<Eliteoomph> - ns2
04:26-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
04:26<rnowak>Peng: not if you're an elephant
04:27<Eliteoomph>lol i couldn't even find there customer service
04:27<dcraig>Eliteoomph, and what's the domain?
04:27<rnowak>Just like that old lost dictionary, I guess
04:27<Eliteoomph>sorry i am bad about that
04:27-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
04:27<Peng>I'm not joking. I know everybody hates GoDaddy, but I've heard they have clueful phone support.
04:27<rnowak>Just in, Peng calls godaddy customer support to get some human interaction
04:28<Peng>rnowak: Why do you think I get so many delivery pizzas?
04:28<Eliteoomph> is one
04:28<Eliteoomph>this one is for
04:28-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
04:29-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
04:29<@heckman> does not have
04:29<Eliteoomph>wish they had live chat... i love live chat support
04:29<Eliteoomph>what is it?
04:29<dcraig>so it won't let you set the nameservers for to and
04:29<@heckman>whois < type that in your Linode's terminal
04:30<Eliteoomph>thats weird
04:30<dcraig>yeah, I hate the number 184, too
04:30<@heckman>Also. Are you seriously hosting your DNS on one machine?
04:30-!-stijlfigurant [] has quit [Quit: CGI:IRC]
04:30<dcraig>two machines in the same rack
04:31<MarkJ>or two IPs on same box, pretty normal for a CPanel setup
04:31<Eliteoomph>what is so bad about that>
04:31<@heckman>Eliteoomph: What happens if the host goes down?
04:31<MarkJ>no redundancy
04:31<@heckman>Eliteoomph: How will your DNS resolve?
04:31<dcraig>well, if you host your nameservers on the same machine as your websites...
04:31<dcraig>if the site is down, do they really need to be able to resolve the hostname? :p
04:31<MarkJ>and to assume all services will be on that machine is generally a bad assumption to make
04:32<Eliteoomph>the justhost nameservers worked fine problem solved ;o)
04:32-!-LittPi [] has joined #linode
04:32<dcraig>linode gives you 5 nameservers in 5 datacenters for free
04:32<Peng>That's the same!
04:32<@heckman>THAT'S TERRIBLE
04:32<Peng>dcraig: Well, free with service.
04:32<rnowak>"added value"
04:33*dcraig adds value to rnowak
04:33<@heckman>Eliteoomph: If you're going to be sticking with Linode...I'd really advise using our Linode DNS Manager. All five of our name serves are in different facilities.
04:33<Zr40>Peng: not the same, last bit differs
04:33<Eliteoomph>thats justhost for yah
04:33*rnowak evolves into mohawk
04:33<MarkJ>well you're paying a provider $2.95/month, guess you get what you pay for
04:33-!-hipsterslapfight [] has joined #linode
04:33-!-hipsterslapfight is now known as spkitty
04:33<@heckman>So even if California does split off from the continental US...your records will still be served.
04:34<rnowak>what if dallas goes under?!
04:34<MarkJ>umm wtf. the JustHost home page has different prices to the prices on the "Web Hosting" page?
04:34<Eliteoomph>heckman I will.... I use linode for my better accounts.... but I have a reseller account with justhost.... for you crappy ones
04:34<dcraig>once california forms its own country, we'll still probably remain connected to the internet...
04:34<Eliteoomph>I have the linode backup service also
04:34<@heckman>I mean like...fall in to the Pacific...
04:35<dcraig>our underwater civilization will also have internet
04:35<@heckman>"Oh wait, why is California floating away....?"
04:35<Jippi>Hi guys.. I'm looking into a subscribe / publish system for my servers, so I can execute commands easily on all my web nodes (like "clear cache" ect.) been looking into AMQP, but doesnt look like it can provide feedback with the result back to the original publisher.. would like a sub/pub where I can get the result / logs back and store
04:35<@heckman>dcraig: I guess technically you already have fiber down there. FiOS for every one!
04:35<Peng>Zr40: Right. That doesn't count.
04:35<MarkJ>Jippi or even a subscription service like Amazon SNS/SQS?
04:35<Peng>Zr40: Also, I meant that it was the same two IPs as before.
04:35<@heckman>Jippi: Your text was cut off. IRC supports 512 bytes
04:36<@heckman>"logs back and store" -- last we got.
04:36<Jippi>that was all there was :)
04:36<rnowak> #1
04:36<Zr40>heckman: not sure if this is the case for Jippi, but some IRC clients automatically send the line once it reaches 512 bytes
04:36<@heckman>Oh, you were darn close to the 512. My mistake.
04:36<@heckman>Zr40: irssi doesn't by default.
04:36<Zr40>heckman: don't forget the PRIVMSG and #linode and such
04:36*heckman found that out the hard way.
04:36<rnowak> #1 *cough*
04:37<dcraig>well if that was the full message, maybe it could be repeated in english
04:37<Jippi>MarkJ: would like to host it myself, so Amazon is not an option :)
04:37<@heckman>I think the "wth, irssi doesn't split lines" convo happened in here.
04:38-!-xt3mp0r [~xt3mp0r@] has joined #linode
04:38<@ericoc>bleh, wrong window <3
04:38<@ericoc>this isn't ssh
04:38-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
04:39<dcraig>feel free to type shell commands, and we'll try to give appropriate responses
04:39<rnowak>if you guys want proper responses, at least provide the password instead of asterisks, jesus.
04:40-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
04:40<@heckman>MacBook keyboad woke up the GF :< -- to the kitchen I go.
04:40-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
04:40<rnowak>heckman: sure it wasn't your breathing?
04:41<dcraig>you don't need to pound the keys for them to work
04:42<rnowak>^ heckman
04:42-!-pygi [] has quit [Ping timeout: 480 seconds]
04:43<MarkJ>someone has that gif on one of my guild forums
04:43<Peng>Wow. That animation just got less pleasant as it went on...
04:43<MarkJ>as avatar
04:44-!-djg320 [] has joined #linode
04:46<Eliteoomph>reminded me of joe cartoon
04:47*heckman is relocated to kitchen
04:48<@heckman>I wasn't pounding the keys, she's just a light sleeper.
04:48<Eliteoomph>but that took a while... what kinda hardware you packing over there?
04:48<@heckman>I sneezed once and she almost hit the ceiling.
04:48<@heckman>Eliteoomph: I also proceeded to go to the bathroom and put some pants on.
04:48<Eliteoomph>have yall ever seen a joe cartoon... cartoon?
04:48<Eliteoomph>lol k heckman
04:49<Eliteoomph>ran across joe cartoon years ago surfing the web... some are kinda funny some are just plain disturbing... def interesting though
04:49<Eliteoomph>he really hates gerbils for some reason
04:50*heckman raged after 45 seconds
04:55-!-Eliteoomph [] has quit [Quit: CGI:IRC]
04:56-!-mathew [] has joined #linode
04:57-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
04:57<Peng>Augh what. Dallas -> Newark goes Dallas -> San Jose -> Ashburn in SoftLayer's network.
04:58-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
04:58<+linbot>Peng: [mtr] 20 hops, 20.0%/77.4ms
04:59-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
04:59-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
05:02-!-darkbeholder [] has quit [Read error: Connection reset by peer]
05:03-!-spkitty [] has quit [Quit: spkitty]
05:03-!-darkbeholder [] has joined #linode
05:04-!-xt3mp0r [~xt3mp0r@] has joined #linode
05:05-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
05:06-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
05:06-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
05:07-!-hipsterslapfight [] has joined #linode
05:07-!-hipsterslapfight is now known as spkitty
05:09-!-spkitty [] has quit [Read error: Connection reset by peer]
05:10-!-hipsterslapfight [] has joined #linode
05:10-!-hipsterslapfight is now known as spkitty
05:12-!-xt3mp0r_ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
05:12-!-xt3mp0r [~xt3mp0r@] has quit [Read error: Connection reset by peer]
05:14-!-xt3mp0r [~xt3mp0r@] has joined #linode
05:15-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
05:15-!-atula [] has quit [Ping timeout: 480 seconds]
05:20-!-hipsterslapfight [] has joined #linode
05:21-!-burningdog [] has joined #linode
05:22-!-LittPi [] has quit [Quit: CGI:IRC (Ping timeout)]
05:25-!-spkitty [] has quit [Ping timeout: 480 seconds]
05:28<aussiekid>what's reverse mapping attempt mean?
05:29<rnowak>I hope you appreciate that link, it required effort and energy to no end to google it.
05:32<@Praefectus>what link?
05:32*rnowak stares at Praefectus
05:34<Zr40>I don't see how a DNS PTR mismatch constitutes a "POSSIBLE BREAK-IN ATTEMPT!"
05:35<+linbot>New news from forums: Hosted Desktops in General Discussion <>
05:35<rnowak>PTR mismatching hosts should not be allowed on the interweb
05:35<Zr40>but it does happen for various reasons
05:35<Zr40>negligence, multiple host names for a single IP, to name a few
05:36<rnowak>the first one is really the issue, the second is just sillyness on their part
05:37*Zr40 intentionally breaks his PTR so that he can now break in to ssh daemons!
05:37<aussiekid>ok thanks will have a read
05:37<aussiekid>oh ive actually read that link
05:37<rnowak>Zr40 such a hax0r
05:37-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
05:38-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
05:39-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
05:40-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
05:43-!-mcinerney [] has joined #linode
05:43-!-burningdog [] has quit [Quit: burningdog]
05:45<fo0bar> <-- booya (finnix dev build running on Linode via pv-grub, nothing linode-specific)
05:46<fo0bar>in the past, the version of Finnix running on Linode has always been at least slightly customized. linode/finnix-specific kernel, modified initrd, etc
05:47-!-amitz [~amitz@2600:3c01::f03c:91ff:fe93:6752] has quit [Server closed connection]
05:47-!-amitz [~amitz@2600:3c01::f03c:91ff:fe93:6752] has joined #linode
05:48-!-guinea-pig [~orion@2600:3c01::f03c:91ff:fe96:469b] has quit [Server closed connection]
05:48-!-guinea-pig [~orion@2600:3c01::f03c:91ff:fe96:469b] has joined #linode
05:48<Peng>Congrats. :)
05:49-!-array [danny@2600:3c01::f03c:91ff:fe96:db42] has quit [Server closed connection]
05:49-!-array [danny@2600:3c01::f03c:91ff:fe96:db42] has joined #linode
05:49<Peng>I wonder if I'm going to get shed.
05:49<Peng>Also, none of you three have rDNS!?
05:49<Peng>For shame!
05:50-!-Bowbles [] has quit [Server closed connection]
05:50-!-Bowbles [] has joined #linode
05:50<Peng>Like that. :D
05:51-!-Parallax_ [] has joined #linode
05:51-!-alt_ [] has quit [Server closed connection]
05:51-!-alt_ [] has joined #linode
05:52*Zr40 sees a rDNS for this IP
05:53-!-Parallax [] has quit [Ping timeout: 480 seconds]
05:53-!-Parallax_ is now known as Parallax
05:54-!-stafamus [~stafamus@] has joined #linode
05:55-!-sneakums [sneakums@2600:3c01::f03c:91ff:fe96:fa6] has quit [Server closed connection]
05:55-!-sneakums [sneakums@2600:3c01::f03c:91ff:fe96:fa6] has joined #linode
05:56<rnowak>yay moar bettar finnix <3
05:56<rnowak>there's always a disc of finnix in my laptop bag, which is always with me <3
05:57-!-JDLSpeedy [] has quit [Server closed connection]
05:58-!-npmr [] has quit [Server closed connection]
05:58-!-npmr [] has joined #linode
05:58-!-JDLSpeedy [] has joined #linode
06:01-!-Boss [] has quit [Quit: Leaving]
06:06-!-Dreamer3 [] has quit [Quit: Computer has gone to sleep.]
06:06-!-xt3mp0r_1 [~xt3mp0r@] has quit [Read error: Connection reset by peer]
06:06-!-xt3mp0r__ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
06:06-!-djg320 [] has quit [Ping timeout: 480 seconds]
06:10-!-JJPEPO [] has quit [Server closed connection]
06:11-!-xt3mp0r [~xt3mp0r@] has joined #linode
06:11-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
06:12-!-scorche|sh [] has quit [Server closed connection]
06:13-!-scorche|sh [] has joined #linode
06:13-!-Josh [] has quit [Server closed connection]
06:13-!-Josh [] has joined #linode
06:14-!-user9180 [] has quit [Ping timeout: 480 seconds]
06:15-!-disinpho [] has joined #linode
06:16-!-asedeno [] has quit [Server closed connection]
06:16-!-asedeno [] has joined #linode
06:17-!-vandemar [spin@2001:470:1f10:56b::4] has quit [Server closed connection]
06:17-!-vandemar [rings@2001:470:1f10:56b::4] has joined #linode
06:18-!-ajmitch [] has quit [Server closed connection]
06:18-!-ajmitch [] has joined #linode
06:19-!-chesty is now known as dolphin_breath
06:19-!-Badgerer [] has quit [Server closed connection]
06:19-!-Badgerer [] has joined #linode
06:20-!-dolphin_breath is now known as chesty
06:20-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
06:21-!-straterra [] has quit [Server closed connection]
06:21-!-straterra [~straterra@2001:470:8a81::4] has joined #linode
06:22-!-xt3mp0r__ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
06:22-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
06:22-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
06:26-!-xt3mp0r [~xt3mp0r@] has joined #linode
06:26-!-jarryd [] has quit [Server closed connection]
06:26<@heckman>What do I do when OS X doesn't kill it in time?
06:26-!-jarryd [] has joined #linode
06:26<@heckman>KILL DASH NINE!
06:27-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
06:27-!-graq [] has quit [Server closed connection]
06:28-!-graq [] has joined #linode
06:28-!-LittPi [] has joined #linode
06:29-!-verb [~blarg@2600:3c01::f03c:91ff:fe93:da] has quit [Server closed connection]
06:31-!-mr_itk [] has joined #linode
06:32<LittPi>actually I suppose to use mysql_real_escape_string for escaping to prevent sql injection
06:32<LittPi>but I see that wikipedia has said it is error prone, i am wondering why
06:32<mr_itk>Hi! Question - if previously i had my LAMP stack with mod_php, then added mpm_itk module from this tut: How to know what Apache config directives are used (server limits etc.)
06:33<rnowak>you'd want to validate input, so that you're getting just what you want, and ideally use prepared statements
06:33-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
06:33-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
06:34-!-deaton [] has quit [Server closed connection]
06:34-!-deaton [deaton@2600:3c00::f03c:91ff:fe96:248e] has joined #linode
06:34<hawk>LittPi: Use parameterized sql statements
06:35-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
06:35-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
06:35<LittPi>nice site, thank you :)
06:36<hawk>LittPi: Do note that they have actual examples of how to do things right in different environments, beyond the xkcd comic :P
06:36<rnowak>(and you should still do input validation)
06:36-!-jimcooncat [] has joined #linode
06:36<Zr40>rnowak: (but that's unrelated to SQL)
06:37<rnowak>(pretty much related to his use-case)
06:37-!-xt3mp0r__ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
06:37-!-xt3mp0r_1 [~xt3mp0r@] has quit [Read error: Connection reset by peer]
06:38-!-k00pa [] has quit [Server closed connection]
06:38-!-k00pa [] has joined #linode
06:40-!-Kuukunen [] has quit [Server closed connection]
06:40-!-Kuukunen [] has joined #linode
06:40<priyesh>what's the default TTL in linode dns manager?
06:41<@heckman>One hour, I believe.
06:41<chesty>which hour?
06:42-!-xt3mp0r [~xt3mp0r@] has joined #linode
06:42-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
06:42<hawk>86400 s, it appears
06:42-!-shirro [] has quit [Quit: shirro]
06:43<hawk>so, 24h
06:43<@heckman>Apparently I was way wrong.
06:43-!-k00pa [] has left #linode []
06:43-!-cats [] has quit [Server closed connection]
06:44-!-cats [] has joined #linode
06:45<priyesh>heckman: is it possible to use the linode DNS but give clients
06:46<@heckman>priyesh: I guess technically you could...but it may break in the future.
06:46<priyesh>if they just did some $dig they'd be able to see it's linode i guess
06:47<priyesh>so i should copy all my dns records over to linode and then change it to with my registrar
06:47<hawk>priyesh: So a thin layer of obfuscation is in order?
06:47<priyesh>for no downtime
06:47<priyesh>hawk: yeah
06:48-!-dominikh [] has quit [Server closed connection]
06:48-!-dominikh [] has joined #linode
06:49<hawk>priyesh: Yes, set your zone up at linode first, then change the nameservers with your registrar to ns{1,2,3,4,5}
06:49<@heckman>priyesh: You could set 'A' records, but if we change our IP addresses you're pretty much hosed.
06:49<@heckman>Not saying it would happen, but you never know.
06:49<priyesh>heckman: yeah. don't want to risk losing website + email
06:50<@heckman>May I pose an inquiry. Why must the fact the DNS is being handled be obscured?
06:50<hawk>Also, as AAAA records appear you won't get them, etc
06:50-!-AviMarcus [~avi@] has joined #linode
06:50<priyesh>heckman: would it work if i set CNAMEs in my main domain? i think DNS doesn't allow CNAMES for NSs though..
06:51<@heckman>priyesh: Correct.
06:51<priyesh>heckman: not to worry, i'm going to stick with ns(1-5) just to be safe
06:51<@heckman>Sorry, correct that it won't work.
06:51<hawk>priyesh: No, just stick to the KISS principle and go with the nameserver owners name for the nameservers
06:51-!-JediMaster [] has joined #linode
06:51<priyesh>hawk: heckman: thanks
06:52<JediMaster>hi all, I've finally got my awesome file system clustering working with DRBD and GFS2!
06:52<JediMaster>on linodes of course =)
06:52<JediMaster>I've done load balancing in the past by using HAProxy but with apache back-ends
06:52<@heckman>Only took you like two weeks... kidding! :p
06:52<JediMaster>10 days ;-)
06:53<rnowak>congratulations \o/
06:53<JediMaster>anyhow, I'm using nginx for everything now
06:53<priyesh>what kinds of things would you use SRV records for?
06:53<JediMaster>and was wondering if I should use nginx for the load banalcing insted of haproxy, what do you guys think?
06:53<@heckman>Not going to lie, Nginx is pretty sexy.
06:53<JediMaster>it is
06:53<hawk>JediMaster: How does it perform?
06:53<JediMaster>but not used it for load balancing
06:54<@heckman>Admittedly, I am not to sure how Nginx fares against HAProxy. Could always be lazy and wait for NodeBalancer to go stable. :p
06:54<mr_itk>no mpm_itk users...?
06:54<rnowak>it performs excellently, but I can't say if it works better than haproxy
06:54<hawk>priyesh: SIP, Kerberos, etc comes to mind
06:54<JediMaster>hawk: well... 2 servers in teh cluster, had to set it up using unicast rather than multicast... took down server 2, copied a 1GB file across partitions to the GFS2 partition, fired up server 2, and within 30 seconds it's and synced up
06:54<@heckman>mr_itk: The activity here fluctuates. You may want to throw a thread on the forums.
06:54<AviMarcus>priyesh, VoIP certainly
06:54<+linbot>heckman: Hosted Desktops in General Discussion <> || network external storage in Sales Questions and Answers <> || CPU Spike in General Discussion <> || looking for someone who knows the API in General Discussion <> || [ Poll ] Linode (1 more message)
06:54<JediMaster>heckman: what does nodebalancer use?
06:55*tonyyarusso wishes the Linode DNS manager had lots and lots more options for record types - someday...
06:55<AviMarcus>iptables stuff?
06:55<rnowak>tonyyarusso: like?
06:55<@heckman>JediMaster: Admittedly, no idea on specifics. I know it uses an off the shelf software plus some baked on goodies.
06:55<JediMaster>heckman: also any ideas at all on price? is it likely to cost more than the price of a 512 'node?
06:56<tonyyarusso>rnowak: Like the ones for storing GPG public keys, mail spam prevention rules, etc.
06:56-!-mattg_ [] has quit [Server closed connection]
06:56<JediMaster>heckman: and, does it give shiny stats showing up and down servers etc like haproxy does?
06:56<rnowak>tonyyarusso: you mean TXT records?
06:56<@heckman>JediMaster: You like asking questions I don't know the answers to. :p -- Price is still being discussed internally.
06:56-!-mattg_ [] has joined #linode
06:56<tonyyarusso>rnowak: No.
06:56<JediMaster>heckman, yes, yes I do =)
06:56<rnowak>tonyyarusso: what record types are those?
06:57<@heckman>JediMaster: If you are curious you can play with it while it's still beta. :p
06:57<JediMaster>heckman: I suppose I can, no reason why I can't have two load-balancers working on the same backend servers
06:57<JediMaster>heckman: where's the signup?
06:58<AviMarcus>it's only in newark datacenter afaik
06:58<Zr40>it's in all five
06:58<@heckman>AviMarcus: wrong
06:58<AviMarcus>oh it changed? cool.
06:58<@heckman>JediMaster: Open a support ticket and ask.
06:58<JediMaster>kk, do I need to give details of backend servers?
06:58<JediMaster>heckman: also does it do ssl load banalcing/failover?
06:58<@heckman>You configure it all in the Linode Manager.
06:59<@heckman>JediMaster: your backends need to terminal SSL, but yes.
06:59<tonyyarusso>rnowak: CERT, LOC, PTR, SPF, and SSHFP records.
06:59<JediMaster>heckman: ok, so just copy the same SSL private and public certs over to each backend server?
06:59-!-shirro [] has joined #linode
06:59<Zr40>tonyyarusso: SPF uses a TXT record
06:59<@heckman>JediMaster: pretty much.
06:59<Zr40>tonyyarusso: I believe LOC does as well
07:00<rnowak>(as does publishing your gpg key)
07:00<rnowak>they appear to be adding random new types for what TXT can already do
07:00-!-flowbee [] has quit [Read error: Connection reset by peer]
07:00<hawk>Zr40: There is actually an SPF rrtype as well, but it works the same as TXT
07:00<tonyyarusso>Zr40: You can do it that way, but that method is less desireable and basically obsolete. Since we have a real record type now we should stop using the legacy workaround.
07:02<hawk>tonyyarusso: Actually I think at this stage the official word is that we must keep the TXT for compatibility as we add the SPF records
07:02<tonyyarusso>DNS lookup tools have a "type" flag. If everything is just chunked together under TXT records that becomes a lot less useful, not to mention the semantic issues.
07:02<tonyyarusso>hawk: Right, because lots and lots of things (like the Linode manager) still lack proper SPF handling.
07:02<tonyyarusso>That's not an argument against implementing the new stuff though.
07:03<Zr40>tonyyarusso: even if you add a SPF record, there are still tons of libs and apps out there that will only look at TXT (and unfortunately they will be for a long time...)
07:03<rnowak>lets add 2000 record types because we can, and not make good use of the old ones already there, like SRV and TXT
07:03<hawk>No, it was an argument against "we should stop using the legacy workaround"
07:04*tonyyarusso sighs
07:04<Zr40>rnowak: problem with TXT is that it's not used for a single purpose
07:04<hawk>rnowak: Well, SRV is one thing... TXT is not used properly for all these special things
07:04<Zr40>rnowak: SRV is indeed underused.
07:05<JediMaster>guys, can I check my logic here, am I right in thinking to get better performance we'd be better off with 2x4GB load-balanced servers than 1x8GB server because there will be effectively twice the cpu power
07:05<@heckman>I'd think so.
07:05<AviMarcus>are you hitting CPU limits?
07:05<AviMarcus>load balance >4 ?
07:05<rnowak>JediMaster: what if your single loadbalancer goes down?
07:05<tonyyarusso>hawk: Allow me to amend that to "should stop *exclusively* using the legacy workaround". You're right otherwise.
07:06<JediMaster>rnowak: there will be two with shared ip
07:06<tonyyarusso>Technically, the RFC says "An SPF-compliant domain name SHOULD have SPF records of both RR types. A compliant domain name MUST have a record of at least one type."
07:06<rnowak>JediMaster: ah, you've got that covered, right
07:06<JediMaster>rnowak: or nodebalancer
07:06<rnowak>oh, load-balanced servers, I read that wrong - sorry :p
07:06<@ericoc>JediMaster: all set, you might need to log out/back in to the manager for the tab to show up
07:07<JediMaster>ericoc: thanks for the super fast response =)
07:07<JediMaster>nice to see you guys are still hot on replying to tickets =)
07:07<@heckman>ericoc is definitely hot. :x
07:07<bob2>pffft 114 seconds
07:07<@ericoc>heckman: <3
07:07-!-mr_itk [] has quit [Quit: CGI:IRC (Ping timeout)]
07:07<@heckman>ericoc: <3
07:07<rnowak>bob2: unacceptable, right?
07:08<Zr40>everyone knows 113 seconds is just barely acceptable
07:08<AviMarcus>JediMaster, what is your cpu usage / load balance?
07:08-!-takamichi [] has joined #linode
07:08<JediMaster>ericoc / heckman don't suppose you guys have a guesstimate on eta for stable release and/or price? nothing official
07:09<@ericoc>nothing yet :\
07:09<tonyyarusso>Not to mention that DNSSEC support should happen soon.
07:09<JediMaster>woo, I can see nodebalancers after relogging =)
07:10<bob2>he dnnsec
07:10<hawk>tonyyarusso: You mean for the manager?
07:10<JediMaster>I see, so you're charging by number of concurrent connections then?
07:10<JediMaster>or max block of
07:10<tonyyarusso>hawk: yes
07:11<hawk>tonyyarusso: The Linode nameservers already do dnssec at least (for slave zones), so it's partially there at least :)
07:12<JediMaster>ericoc: can I use internal IPs?
07:13<tonyyarusso>hawk: Yeah, they'll mirror the other record types if you run your own DNS master too - it's only the Manager interface that lacks stuff.
07:14<hawk>tonyyarusso: I didn't mean them not barfing on the rrtypes, they do the actual dnssec bit (it's not just serving the records, there is some special handling)
07:14<Peng>JediMaster: Using two 4 GB nodes gives you much more transfer.
07:14<Peng>JediMaster: Since a 4 GB provides 1600 GB of transfer, but the larger plans level off at 2000 GB.
07:14<@heckman>JediMaster: internal IPs for...?
07:14-!-jimcooncat [] has quit [Quit: Ex-Chat]
07:15<JediMaster>Peng: true, forgot about that too
07:16<JediMaster>heckman: answered my own question, nodebalancer automatically fills in my internal ips
07:16<Zr40>I wonder if nodebalancer could work on with any public IP as backend
07:16-!-flowbee [] has joined #linode
07:16<JediMaster>heckman: however... I have it setup to point at two of the server's internal IPs, and it's got HTTP valid status check enabled, but the status of both nodes is unknown/down
07:16<bob2>Zr40: no
07:17<JediMaster>heckman: and I can see the requests coming into nginx from
07:17<JediMaster>heckman: with a 200 response
07:17<bob2>Zr40: must use private linode ipv4 address
07:17-!-tiny [] has joined #linode
07:17-!-xt3mp0r [~xt3mp0r@] has quit [Read error: Connection reset by peer]
07:17-!-xt3mp0r_ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
07:17<bob2>i think it supported v6 backend addresses initially
07:18<Zr40>according to it still does (it mentions native IPv6)
07:18<bob2>the /frontend/ sure has v6 native
07:19-!-xt3mp0r [~xt3mp0r@] has joined #linode
07:19-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
07:19<@heckman>JediMaster: Looks like it might be a GUI issue. It looks like your NodeBalancer's IP is serving content.
07:20<Zr40>nodebalancer does not work across DCs, does it?
07:21<Peng>JediMaster & heckman: I've heard of a bug like that.
07:21<JediMaster>heckman: ah ok, I've just reported it in the support ticket
07:21<@heckman>Zr40: I *believe* it can...
07:21<Peng>19:50:31 < jed> the UI isn't updated due to an unrelated bug that I'm fixing today
07:21<@heckman>Zr40: But I am not 100% sure.
07:21-!-atan [] has quit [Read error: Operation timed out]
07:21<Peng>It's been fixed, but may not have been rolled out yet.
07:22<bob2>Zr40: pretty sure it needs a local private ipv4 address
07:22<bob2>no way does linode have < a /16 of nodes
07:22<JediMaster>heckman: here's a nasty question, can the nodebalancer's IP failover to a different DC should there be network trouble with the DC?
07:23<@heckman>IPs cannot failover in any situation.
07:23<@heckman>Well, to another DC...
07:23<bob2>JediMaster: effectively impossible
07:24<JediMaster>I thought there was a way of redirecting ips or blocks of ips at any rate to entirely different countries if needed?
07:24<bob2>unless the nodebaancer ips are put in /24s that are reannounced from another dc
07:24<JediMaster>that's the one
07:24<bob2>i don't know that linode controls their routing sufficiently everywhere to do that
07:24<bob2>and it would mean /everyone/ in that /24 gets shifted
07:24<bob2>if you're that srsbzns get your own ip blocks (quickly) and racks
07:25<rnowak>bob2: if a different router announces for a block, would it effectively pick it up immediately, or are there bad things that could happen?
07:25<bob2>rnowak: any number of routers can announce routes (anycast is when you deliberately do that)
07:25<rnowak>I see
07:26<bob2>rnowak: your traffic will end up at whichever 'endpoint' they think is closest
07:26<rnowak>they being the routers the source destination is using?
07:26<JediMaster>bob2: if it was for nodebalancers only and the entire DC had network issues it might work
07:26<rnowak>or rather, the routers on the path
07:26<bob2>it's hard to detect even, since if I announce one block from one dc via multiple ISPs, you might see that announcement from more than one of those
07:26-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
07:27<JediMaster>meh, office temperature has climbed 3 degrees over the last 3 hours =/
07:27<bob2>rnowak: depends on the traffic direction (the directions might go totally different paths)
07:27<rnowak>bob2: right
07:27<hawk>rnowak: "they" being any router in the path
07:27<bob2>JediMaster: if everyone in that dc on that nodebalancer agreed on a fallback dc
07:27-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
07:27<hawk>rnowak: and "closest" being least number of ASes to pass through
07:27<bob2>^ subject to router admin fiddling
07:28<rnowak>bob2: so with a funny router, A to B, all routers between could think B is closest, but then all of a sudden the last decides C is closer, and goes there?
07:28<rnowak>or no such funny things can happen?
07:28<Zr40>yes, that's possible
07:28-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
07:28<bob2>dnsmadeeasy uses the above to do global load balancing
07:28<bob2>announce addresses from lots of places
07:28<Peng>Lots of companies do that.
07:28<hawk>rnowak: Funny stuff can happen, and the more creative you get with your BGP announcements the higher the risk would seem to get
07:29<bob2>when the AU node gets traffic, assume it came from an AU client, return an A record saying to use an AU web server
07:29-!-JSharp [] has joined #linode
07:31<rnowak>I could probably read it up somewhere... but while we're on the subject ;p So when a router announces, all other routers on the internet pick it up and store it, or is there some querying going on?
07:31<Zr40>rnowak: read up on BGP
07:31-!-xt3mp0r [~xt3mp0r@] has joined #linode
07:31<hawk>rnowak: Routers announce what prefixes they know, and the path to them to their _neighbours_
07:31<bob2>so, 'real' routers have no default route
07:32<bob2>they talk to their neighbours via BGP and keep a full routing table for the internet
07:32<hawk>bob2: default routes are more of a client computer thing
07:32<bob2>when they get traffic to route, the look up the routing table, find the route with the fewest ASNs (subject to config fiddling) and spit it at the router that told it about the route
07:33<hawk>bob2: (I think in general the fiddling is to actually modify the number of ASNs, ie, inject a bunch of the same ASN in the path to make it longer)
07:33-!-Zr40 [] has quit [Quit: leaving]
07:33-!-Zr40 [] has joined #linode
07:34<JediMaster>don't suppose anyone knows how to replace the remote ip with x-forwarded-for ip in nginx logs?
07:34<bob2>prepending right?
07:34<hawk>bob2: Yeah, I guess
07:36<hawk>Specifically "The headers, transmitted to client, begin from the prefix "sent_http_", for example, $sent_http_content_range", I'd think
07:36<hawk>Or was this the ot her way around?
07:36<hawk>Maybe I'm confused
07:37<Peng>Remember, year or two back, when someone accidentally prepended an ASN hundreds of times and crashed a bunch of routers? :D
07:37<hawk>Peng: Yep
07:37<rnowak>hmmm, is there a risk with ipv6 that the routing tables will explode, or are the poor routers prepared to cope with it?
07:37<@ericoc>i've heard a few stories about single routers being tweaked to stop routing youtube and it affecting entire continents/multiple countries
07:38<Peng>(They neglected to RTFM and set, like, "prepend = 250 250", expecting it to prepend 250 twice. Instead, it prepended it 250 times.)
07:38<rnowak>looking at information about the BGP tables now
07:38<Peng>ericoc: Yeah, Pakistan did that.
07:38<@heckman>Yeah. Failure at pushing that change via BGP. I think I read something once about someone pushed a change via BGP and a large chunk of traffic was routed through Turkey...
07:38<@heckman>As it was "the best route"
07:39<@ericoc> - but thats older
07:39<hawk>rnowak: The prefixes that are advertised are quite large for v6, but yes, that is a challenge
07:39-!-mikeage [] has joined #linode
07:39<rnowak>I see
07:39<bob2>rnowak: they're trying to allocate v6 so it won't be such an issue
07:39<Peng>rnowak: Well, it's complicated. Ultimately, since major ISPs will have one /29 or whaever instead of a dozen small IPv4 allocations, the IPv6 routing table will be smaller -- or so people think/hope. For now though, with routers having to handle both the full v4 and v6 tables, yes, size is a problem.
07:40<+linbot>Bigger is better.
07:41<Peng>Follow for global routing incidents. :)
07:41<+linbot>THAT'S WHAT SHE SAID!
07:41<bob2><3 linbot
07:41*ericoc adds to google reader
07:41<AviMarcus>wow, so one malicious router operator can screw up nearly the whole 'net?
07:42<Peng>AviMarcus: Nowadays that router operator would have to be at a major ISP, but yes.
07:42<bob2>Peng: oh, btw, akamai isn't offering their whole cdn via v6 yet, that's why it has no AAAA records
07:43<Peng>AviMarcus: The number of routes it would require to disrupt the entire Internet would very quickly be noticed, though.
07:43<AviMarcus>I see.
07:43<Peng>AviMarcus: Taking down YouTube, like Pakistan accidentally did? No problem. But the entire Internet would be trickier/.
07:43<@heckman>Peng: You ever seen the series Red vs. Blue?
07:43<Peng>bob2: Wait, what's the context of that? Were we talking about Akamai?
07:43<Peng>heckman: No.
07:44<@heckman>Peng: Lamesauce :<
07:44<@heckman>What you said brought a quote to mind from it, hah
07:44<@ericoc>didn't some south american country purposely try to prevent their citizens from visiting youtube with BGP, but the bogus route sort of accidentally "spread"
07:44<bob2>Peng: re: cloudfiles not being available v6 yet (the cloudfiles thing itself is, just not via the CDN)
07:44<Solver>that was pakistan
07:44-!-jonsowman [] has quit [Server closed connection]
07:44-!-jonsowman [] has joined #linode
07:45<JediMaster>heckman or ericoc: how long does it take to kick in with a new setting for the balancing type? changed it about 15 min ago from source ip back to round-robin and it's still stuck on source ip but shows round robin in the interface
07:45<bob2> <- deets
07:45<@ericoc>JediMaster: no clue, i'm not the expert on nodebalancer either, maybe ask in #linode-beta
07:46<Peng>bob2: Oh oh right right.
07:46<JediMaster>will do
07:46<bob2>!blame jed
07:46<@ericoc>^ that too
07:46<rnowak>I take it there's some mechanisms to prevent random people announcing for blocks they don't "own" - would that be the responsibility of the neighbouring routers, only allowing ones they know they should be routing for, and if so, who decides that? Is there some docu you could recommend for reading about this? It is a bit of black magic to me ;p
07:47<Peng>Did you follow the link to the NANOG post about the AS7007 incident? It ends:
07:47<Peng>"I would also like to take this time to thank AT&T WorldNet, NASA Sciences Institute, and Net Access Corporation who called and did not just ask for an explanation, but offered asisstance."
07:47<Peng>*The* Net Access Corporation? :D
07:47<JediMaster>I wouldn't do that to jed, he spent an hour yesterday trying to help me with OCFS2, before we both gave up, I then tried GFS2 and got it working in 5 min
07:47<rnowak>The Interweb Inc.
07:47<rnowak>ericoc: thanks
07:47<JediMaster>heh, what have I started with all this load balancing talk? =)
07:47<rainman_>there's some techniques for routing filtering
07:47<rainman_>but none are fool-proof, as practice shows all the time
07:47<Peng>Oh, NAC's website says they've been around since 1995. I hadn't noticed.
07:47<bob2>nac used up all their clue on that incident, thus filtering irc
07:48<rnowak>JediMaster: you've given me something to do, thanks ;p
07:48<Peng>JediMaster: Check out #linode-beta
07:48<Peng>bob2: That's Atlanta, not NAC.
07:48<bob2>is newark nac?
07:48<Peng>bob2: Yes
07:48<@Praefectus>atlanta is gnax
07:49-!-Edgeman [] has joined #linode
07:49<bob2>f: he, a: gnax, d: softlayer, l: telecity?
07:49<Peng>bob2: Yes
07:49<bob2>oh, there's a wiki now?
07:49<bob2>who knew
07:49<Peng>The wiki's been around forever.
07:49<rnowak>that wiki appears to be ancient
07:50<bob2>zomg :/
07:50<Peng>It predates the library and most of the content is out-of-date and not very good anyway.
07:50-!-mikeage [] has left #linode [Leaving.]
07:50<Solver>it's alittle known fact that Neanderthals used wikis extensively
07:50<bob2>Solver: that's an absurd liberal myth
07:51<Peng>Eek, both GNAX and NAC's websites have nameservers on sequential IPs. :X
07:51<bob2>Solver: they used oil from the arctic
07:51<Solver>I resemble that remark! :)
07:51<rnowak>the cavemen had to collaborate in their hunting endavours, obviously
07:51<Peng>NAC's are dual-stack, but still sequential.
07:51<Solver>quite right. it was infact their continued reliance on IPv4 that resulted in their eventual extinction
07:51<hawk>Peng: s/websites/domains/
07:52<Solver>let us not make the same mistake!
07:52<Peng>hawk: The domains used by their websites.
07:52<hawk>Peng: And used by other stuff as well
07:52<Peng>hawk: Probably, but I don't know that.
07:53<hawk>Peng: I would think those internet dudes have email and whatnot
07:53<rnowak>emails, pff, modernities
07:53<bob2>haha is squatted
07:54<rnowak>was it ever's?
07:54<Peng>bob2: You should have warned me it was GoDaddy! Eek
07:54<hawk>rnowak: Probably not
07:54<bob2>rnowak: registered in 1992
07:54<bob2>so maybe
07:54<AviMarcus>what about godaddy?
07:54<bob2>AviMarcus: terrible
07:55<Peng>AviMarcus: is a GoDaddy parking page.
07:55<AviMarcus>price? sell-ups? DNS records?
07:55<Peng>...What about DNS records?
07:55<bob2>terrible UI
07:55<rnowak>terrible name
07:55<AviMarcus>that page is completely botched in chrome
07:55<bob2>that said, their dns features are pretty good
07:55<bob2>v6 glue
07:56<hawk> reveals that used to be owned by "NetAbility Corp"
07:56<@heckman><3 NAC
07:56<Peng>!dns6 ns
07:56<+linbot>Peng: 2001:518::1,
07:56<+linbot>Peng: 2001:518::2,
07:56<Peng>I cannot love anyone who would do that!
07:56<Peng></3 NAC
07:57-!-Edgeman [] has quit [Ping timeout: 480 seconds]
07:57-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
07:57<Peng>!dns6 a
07:57<+linbot>Peng: The query name does not exist.
07:57<Peng>Bad Linode! :O
07:58<@heckman>softlayer isn't sequential...but it's close enough.
07:58-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
07:58*heckman teases Peng
07:58<bob2>maybe they anycast the above names
07:58<bob2>er ips
07:58<bob2>what could possibly go wrong
07:59<Peng>bob2: That completely fails if the route gets backhoed, right?
07:59<Peng>It doesn't? Where do the packets go?
07:59<bob2>then the surrounding routers pull the route and your traffic goes to somewhere else
07:59-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
07:59<bob2>^ yes this is not instantenous
07:59<Solver>anycasting is announcement of the same network ffrom multiple locations
07:59<Peng>if it's not instantaneous, that's "completely fails"
07:59<Peng>A bit.
07:59-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
08:00<Peng>It's bad enough, anyway.
08:00<rnowak>fails a bit completely
08:00<Solver>GP uses a simple shortest path algorithm
08:00<Solver>it knmows of multiple routes and simply uses the shortest path
08:00<Solver>no network convergence is required
08:00<bob2>I don't know how quickly they notice backhoage
08:00<Solver>BGP updates are fast
08:01<bob2>it needs to notice it went down
08:01<Solver>the bgp routers receive route updates
08:02<Solver>if a tcp connection loses some packets it will resend
08:02<Solver>the real problem is state :)
08:02<Solver>anycastnig doesn't work for every application
08:02<bob2>won't it at least take TCP timeout for the route to get pulled then?
08:02<Peng>We were talking about DNS here.
08:02<Peng>Which is largely UDP.
08:03<Solver>the resolver will retry if it loses a connection
08:03<Solver>we have anycasted dnsservers at work.
08:03<Peng>Not much good if there's only one connection.
08:03<Solver>anycasting implies multiple connections
08:04<Peng>But packets don't magically go to every connection. The router chooses one.
08:04<Solver>well multiple sites anyway
08:04<Peng>'s nameservers are on two neighboring /24s for IPv4, but traceroute goes to the same place. IPv6 is all the same /64.
08:04<Solver>shortest path
08:04<Peng>So anycasting N IPs to the same location is no better than having 1 IP.
08:04<Solver>yes packets may be lost but various other layers deal with that quite well
08:04<bob2>Peng: yes
08:04<Solver>Peng: you can anycast to the same location but it may not mak much sense
08:04<bob2>sorry, was joking above, the sequential ips are stupid even if it is anycasted
08:05<@heckman>Apparently is down again..
08:05<Peng>Solver: And yet people do it.
08:05<Solver>Peng: it will buy you some redundancy, eg external links
08:05<Solver>but you wouldn't get the benefit of anycasting across multiple sites
08:06<Solver>but anycastign across sites may be difficult if you need tokeep state between the sites. such things are possible.
08:07-!-vegarl [] has quit [Server closed connection]
08:07-!-vegarl [] has joined #linode
08:09<Solver>Peng: ah i see you mentioned orgs putting their nameservers too close together. That's a whole class of FAIL right there :)
08:09<straterra>my companies nameservers are two vms sitting next to each other in a datacenter
08:09<Peng>straterra: ...Are the VMs both on the same host? :D
08:10<Solver>have you pointed out the problems inherent in this?
08:10<straterra>if they go down, it doesnt matter because the services being provided via dns will also be down
08:10*Solver hands straterra a clue stick to hep deal with the problem
08:10<bob2>hey, if you host your mail and web in the same rack it doesn't matter that much
08:10<straterra>Peng: nah...different hosts
08:10<Peng>straterra: aww
08:10*JediMaster ponders using nginx or nodebalancer to balance mysql connections across two servers in tcp mode
08:10<Solver>straterra: it's worse tolose your nameservers. it does matter
08:10<Solver>that's the mistake a lot of peoplename
08:10<straterra>how is it worse?
08:11<Solver>resolves will get NXDOMAIN
08:11<Solver>mail won't queue, it will bounce
08:11<Peng>They'll get SERVFAIL
08:11<hawk>Solver: No, they will not be NXDOMAIN
08:11<Solver>if all the nameservers are down, try it
08:11<straterra>we use messagelabs....mail will get queued
08:11<bob2>NXDOMAIN requires servers to be up and saying 'no dice'
08:12<hawk>Solver: I can't be bothered, but it won't be NXDOMAIN. NXDOMAIN is something a server has to respond with ("I know this does not exist")
08:12-!-cao [] has joined #linode
08:12<Peng>I think we've already established this, but I want to go over it again: Say you have two authoritative nameserver IPs, on different /24s, anycasted to the same set of locations. A client makes a query to one of the IPs, the query tries to get there but gets killed along the way by a backhoe. The client retries with the other IP. It would take the same route and die in the same place, no? Which means, for some amount of time...
08:12<Peng>...before the routes adjust, you'll be offline, no? Which means it's stupid to do this, right?
08:12<Peng> ++ !
08:12<swaj>heckman: my mom works for mastercard... in their transaction HQ. I should call her and tell her the site is down... though she probably knows :P
08:12<Solver>straterra: most MTAs will bounce mail in that situation. why would you queue mail for a non-existant domain?
08:12<bob2>Solver: no they don't
08:12<straterra>i dont queue the mail...messagelabs does
08:12<Solver>bob2: I've seen it with several MTAs
08:12<bob2>Solver: the recursive resolver will say servfail, since it walked from . and got stuck at the last hop
08:13<straterra>they hold it until the email server is back online
08:13-!-cao [] has left #linode []
08:13<bob2>messagelabs won't be holding anything in the above case
08:13<bob2>originating MTAs will be queing it
08:13<Solver>bob2: that's what I was saying
08:13<Solver>the mail will not queue
08:13<hawk>straterra: How would someone find messagelabs if your nameserver is down? Or in what direction do you mean that they are involved?
08:13-!-blindwaves [] has joined #linode
08:14<bob2>Solver: the mail queues on all the ISP's and google's MTAs
08:14<bob2>Solver: since they can't find the MTA yet
08:14<straterra>hawk: by having an entry at the registrar pointing to messagelabs dnsservers
08:14<straterra>as well as our own
08:14<hawk>straterra: At the registrar? Uh?
08:14<bob2>Solver: consider this - how can the MTA differentiate the above from it having no working local dns servers
08:14<Solver>bob2: didn't you just agree with me above? :)
08:15<straterra>yes, at the know where you put your nameservers?
08:15<hawk>straterra: Yes
08:15<Solver>abou mail queueing in the absence of any reachable nameserver
08:15<bob2>Solver: nothing bounces unless straterra doesn't fix it in 5 days
08:15<straterra>you toss messagelab's in to the mix
08:15<hawk>straterra: So messagelabs are hosting a slave(?) NS for you?
08:15<hawk>straterra: Or what are you saying?
08:16<hawk>straterra: Kinda?
08:16<rnowak>me no follow
08:16<swaj>heckman: my mom says Weds. nights are production push nights. So there was likely a prod push last night, and the web admins are morons, so they didn't keep a rollback backup just in case.
08:16<straterra>they forward all non mx to our nameservers..but thats fine as the mx points to them anyway
08:17<hawk>straterra: That sounds less efficient than having a full slave zoe
08:17*straterra shrugs
08:18<Peng>That's weird.
08:18<Peng>Like, forward? In realtime?
08:18<straterra>with messagelabs, people can log in still to send/receive email too
08:18<straterra>its spiffy
08:18-!-newbie [] has joined #linode
08:19<straterra>anyway, shower time
08:21-!-atula [] has joined #linode
08:21<hawk>I find it curious why they would do some specialized solution with custom software that seems highly inefficient at a first glance.
08:22<newbie>hi all..was hoping to report back with good news from having another crack at going from Unbuntu 9.04 to 10.04 but alas something strange has happened. Got all the way through to rebooting the system after updating from 9.04 to 9.10 then I couldn't SSH back in. I LISHed in and noticed "mountall:/proc: unable to mount: Device or resource busy"
08:23-!-Tiven_ [] has quit [Server closed connection]
08:23<newbie>any harm in issuing a reboot from Linode manager?
08:23-!-Tiven [] has joined #linode
08:26<bob2>prolly have no choice if lish doesn't give you a login prompt
08:27<newbie>bob2: hi again..thanks for is asking "Give root password for maintenance" but it isn't taking my password. glad at least the problem is "weird" and not just "newbie"
08:27<@ericoc>9.10 usually requires that automount devtmpfs be disabled in config profile. you might need to change that and reboot
08:28<@ericoc>anything after 9.10 will need automount devtmpfs on
08:29<bob2>i couldn't find anything about that in 9.10 announcement blog post so gave up
08:29<newbie>the stranger thing still is the fact that I accidentally interrupted by last attempt of going from 9.04 to 9.10 and I didn't get this error. Then again..I didn't get to a point where the upgrade wanted to reboot the system
08:29<@ericoc>automount devtmpfs wasn't added to config profiles as an option until recentlyish
08:29<@ericoc>(after 9.10 came out)
08:30<@ericoc>but 9.10's way out of date, so it doesn't come up much
08:30<newbie>ericoc: thanks for thoughts. here's my problem..I actually can't get in to change anything. keeps saying I have an incorrect login
08:30<bob2>newbie: in the linode web thing
08:30<bob2>newbie: and the login: prompt inside lish expects your linux username/password
08:30<newbie>bob2: "linode web thing" as in LISH?
08:30-!-mathew [] has quit [Quit: Leaving]
08:31<@ericoc>9.10 isn't in the list for deployment in the linode manager anymore
08:31<bob2>newbie: no
08:31<@ericoc>linode web thing = linode manager
08:31<bob2>newbie: same page you change the kernel on
08:32<@ericoc>you'll probably want latest 2.6 paravirt too
08:33-!-jarek [] has joined #linode
08:33<newbie>bob2: as in Linode Manager > Rebuild ?
08:34<bob2>dashboard -> edit link next to the "configuration profile" you're running
08:35<newbie>bob2: there supposed to be something about login here?
08:35<jarek>If I choose "Month to month" payment plan and I will fail to pay one or two times (e.g. because my credit card got locked), will be linode accound be deleted?
08:35<Peng>I'm on 8.04...I have automount devtmpfs on...I *think* I've rebooted since I did that...
08:35<bob2>newbie: 22:27:54 <@ericoc> 9.10 usually requires that automount devtmpfs be disabled in config profile. you might need to change that and reboot
08:35<bob2>newbie: that is the thing you do now
08:36<jarek>s/will be linode accound be deleted/will my linode account be deleted
08:36<bob2>jarek: few days grace after the 1st, some more time may be possible if you raise a ticket and let them know
08:36<@ericoc>Peng: 9.10 is really the only distro that barfs with automount devtmpfs on
08:36-!-freshmilk [~work@] has joined #linode
08:36<Peng>ericoc: Huh.
08:36<jarek>bob2: just few days? will my machine be just locked or completely deleted with all data?
08:37<newbie>bob2, erico: AH..cheers..ok, setting "Automount devtmpfs" to no and heading for reboot. cheers.
08:37<Peng>jarek: You'll be shut down after 10-14 days (don't remember which); I dunno how long before you'll be deleted.
08:37<Peng>jarek: As bob2 said, raise a ticket when this happens.
08:38<Peng>jarek: Alternately get a less sucky CC...
08:38<Peng>jarek: By "this happens" I meant when you're unable to pay.
08:38<newbie>and u guys recommend changing Kernel to Latest 2.6 paravirt (bearing in mind I'm going to go from 9.10 to 10.04 LTS)?
08:38-!-mathew [] has joined #linode
08:38-!-LittPi [] has quit [Quit: CGI:IRC (Ping timeout)]
08:39<bob2>I can't remember what 9.10 needs
08:40<newbie>ericoc: so if 9.10 is the only distro that has a problem with "Automount devtmpfs" set to yes..I should change this back to yes after I get to 10.04 LTS?
08:41<newbie>bob2: ok..might leave Kernel at Latest 2.6 Legacy for the moment then
08:45-!-Edgeman [] has joined #linode
08:46-!-integral [] has quit [Server closed connection]
08:46-!-integral [] has joined #linode
08:46-!-LittPi [] has joined #linode
08:46-!-LittPi [] has quit []
08:46-!-blackroot [] has quit [Server closed connection]
08:47-!-blackroot [] has joined #linode
08:47<Peng>newbie: 10.04 might force you to turn automount devtmpfs on.
08:48<@ericoc>newbie: yes, you'll want automount devtmpfs on when you boot 10.04
08:48-!-amoe [] has joined #linode
08:50-!-JediMaster [] has quit [Ping timeout: 480 seconds]
08:50-!-JediMaster [] has joined #linode
08:51-!-Netsplit <-> quits: FloodServ
08:52<encode>ooh, time to start flooding
08:53<newbie>thanks guys but after rebooting..still have a problem..specifically "mountall start/starting", "init: hwclock main process (832) terminated with status 77", init: ureadahead main process (833) terminated with status 5", "mountall: /proc: unable to mount: Device or resource busy", "mountall: /proc/self/mountinfo: No such file or directory, "mountall: root filesystem isn't mounted", "init: mountall main process (836) terminated with status 1", "General erro
08:53<bob2>maybe need a newer kernel
08:54<bob2>or to check the devtmpfs setting
08:54-!-sean_w [] has joined #linode
08:54-!-ChanServ changed the topic of #linode to: Linode Community Support | | Linode turns 8; Disk space +25% -
08:54-!-FloodServ is "FloodServ" on (unknown)
08:54-!-Netsplit over, joins: FloodServ
08:54-!-mode/#linode [-o tparker] by ChanServ
08:54-!-mode/#linode [-o lsabota] by ChanServ
08:54-!-mode/#linode [-o lmatos] by ChanServ
08:54<bob2>why does OFTC sometimes set +R on me multiple times?
08:54<newbie>bob2: devtmpfs definitely on "no" try this with "Kernel Latest 2.6 paravirt" & reboot?
08:54<encode>bob2: in case you forgot your nick was registered
08:55<StevenK>So you can be *sure*.
08:55-!-DephNet[Paul] [~Paul@] has joined #linode
08:55<Peng>bob2: Cuz services just bounced?
08:55<StevenK>That sounds like pirate mode.
08:55<bob2>Peng: ok
08:56<bob2>how can services tell i'm authed aside from my existing +R?
08:56<Peng>bob2: Using CertFP?
08:57-!-Cromulent [] has joined #linode
08:57<bob2>maybe the servers remember that
08:57<rnowak>perhaps socket state
08:57<rnowak>ehum, as in they keep meta data per connection
08:58<JediMaster>anyone know if the setup in mysql 5.5 is the same as previous versions for multi-master replication?
08:58*JediMaster cheers for thunder outside
08:59<JediMaster>finally it's getting cooler again
08:59<swaj>rofl the tor node moron is so naive it's funny.
08:59<rnowak>swaj: link
08:59<bob2>multimaster :/
09:01<swaj>TLDR: He's running a tor exit node. Linode got multiple DMCA/Abuse complaints and told him to fix his exit node. He QQ'ed and told them there was nothing he could do and it wasn't illegal anyway. Linode said "so what? you broke the ToS." And now he's crying on the forums, yet nobody agrees with him.
09:01<bob2>jesus there's a lot of threads about "wah wah i want more storage"
09:01-!-disinpho [] has quit [Quit: disinpho]
09:01*JediMaster ponders getting his camera out and running to get a photo of the cathedral hit by lightning
09:01<rnowak>nice, let's have a read ;D
09:02<swaj>I find it funny that he accuses Linode of being rude and unprofessional, while he was the only one exhibiting that behavior :P
09:03<rnowak>"Though as a general aside, the DMCA can suck it"
09:03<swaj>Linode was nothing but courteous and professional.
09:03-!-lunks [~lunks@] has joined #linode
09:03<swaj>and far more willing to put up with that crap than I would be :P
09:04<rnowak>I'd take the easy way out and not allow tor exit nodes, because this crap will happen and there's no two ways about it
09:04<Solver>alright I stand corrected. I setup a bogus domain with non-existant nameservers. The mail is queuing, not bouncing and dig did indeed report SERVFAIL
09:04<newbie>nice work guys. "Automount devtmpfs" to "No" and Kernel to "Latest 2.6 Paravirt" let me back in. cheers.
09:05<Solver>I'msure I saw different behaviour from sendmail a long time ago but then my memory may bw wrong too
09:05<HoopyCat>Solver: people still use sendmail?!
09:05<rnowak>"I guess I'll just fall back to the official legal statement from the ToR organization", I lol'd
09:05<Solver>HoopyCat: I did until 8 years ago :)
09:05<bob2>Solver: mad props for doing the experiment
09:05<bob2>i was becoming convinced i was wrong
09:05<Solver>the domain is if anyone wants to play
09:06<Peng>swaj: I think Linode's responses could have been better, though.
09:06<Solver>bqto add insult on injury my internal mail server forward the mail to its relay host without doing a dns lookup at all :)
09:06<Solver>bob: add insult on injury my internal mail server forward the mail to its relay host without doing a dns lookup at all :)
09:06<HoopyCat>Solver: postfix is currently masticating
09:06<swaj>Peng: meh, they did well enough imo. Even if the activity wasn'y illegal, it was definitely against the Linode ToS.
09:06<Karrde>why is the thread 5 pages
09:07<@Praefectus>Karrde: too many people trying to fix stupid
09:07<sean_w>So I set up OpenVPN per the libeary instrutions. But, I need to be able to access the internet via the tunnel. How can I configure that?
09:07<bob2>Karrde: wah wah i wanna run a tor node for $20/month
09:07<newbie>off to kick 9.10 to 10.04 LTS...when do I set "Automount devtmpfs" to "Yes"? Should I set that in Linode Manager prior to starting the upgrade?
09:07*Peng rereads the exchange.
09:07<HoopyCat>sean_w: i think the library article has some stuff on NAT
09:07<HoopyCat>Solver: Jun 28 13:07:35 framboise postfix/smtp[18217]: 290B781C4: to=<>, relay=none, delay=73, delays=0.46/0.01/72/0, dsn=4.4.3, status=deferred (Host or domain name not found. Name service error for type=MX: Host not found, try again)
09:08<bob2>having no MX or A will lead to bounces
09:08<bob2>if yer nameservers work
09:08<rnowak>"Tor is network software that helps users to enhance their privacy, security, and safety online" ok
09:08-!-JSharp [] has quit [Quit: Leaving]
09:08-!-tiny [] has quit [Quit: Leaving]
09:09<swaj>it is, but the problem is that it's mostly used to anonymize illegal activity :P
09:09-!-jarek [] has quit [Quit: Leaving]
09:09<rnowak>safety and security? I don't know, arguable - I guess it depends how you define those terms.
09:09<hawk>s/mostly/also/ would be more fair, I think
09:10<HoopyCat>rnowak: which, according to the "We Can't Have Anything Nice" theorem, means that it will be used by abusers
09:10<rnowak>HoopyCat: indeed
09:10<swaj>hawk: I doubt it.
09:10<Peng>OK, I've reread the exchanges. Linode's responses were better than I thought. The first reply or two, and one later one, came across to me as boiler-plate, impersonal, not-necessarily-clueful support-drone-type responses, but the rest were great.
09:11<Peng>I didn't reread TZ's side of the exchange, though.
09:11<swaj>it was mostly rude
09:11<swaj>and confrontational
09:12<Peng>Right. I'm not going to reread it and confirm my impression, but what came across to me the first time was that Linode wasn't fully listening to him.
09:12<rnowak>"Alright, first off I would ask you be a little less rude about this "
09:12<rnowak>haha, what
09:12-!-storrgie [] has joined #linode
09:12<AviMarcus><HoopyCat>: "We Can't Have Anything Nice" theorom - they say that on mythbusters every once in a while as they blow up cars and stuff ;)
09:12<Peng>Not that I *agree* with what he was saying, but I got the impression a lot of his anger was because he felt ignored, and I don't entirely disagree with that.
09:12<HoopyCat>AviMarcus: ha, they musta stole that from me
09:12<rnowak>not only does he have a tinfoil hat, is appears to also be schizophrenic ;p
09:12<Peng>s/what he was saying/his position or attitude/
09:13<hawk>Peng: I got the impression that he was angry because Linode wasn't fighting his privacy battle
09:13<HoopyCat>there's a fundamental difference between prohibiting technologies and prohibiting behaviors
09:13<Peng>Ha, that too.
09:13<swaj>he expects Linode to go to court for him
09:13<swaj>or fight the DMCA notices
09:14<newbie>To kick Ubuntu 9.10 to 10.04 LTS...when do I set "Automount devtmpfs" to "Yes"? Should I set that in Linode Manager prior to starting the upgrade in the CLI?
09:14-!-Cromulent [] has quit [Read error: Connection reset by peer]
09:14<HoopyCat>newbie: i'd probably have to defer to whatever the library article says... been awhile since i've done that upgrade
09:14<storrgie>Is there documentation for how to move my linode from the UK to another datacenter
09:14<Peng>storrgie: There's not much to document.
09:14-!-Cromulent [] has joined #linode
09:14<swaj>storrgie: Just open a support ticket and Linode staff will migrate it for you.
09:14<@ericoc>storrgie: you just need to open a ticket. requires an ip change and 10-15 minutes per GB of data to be moved
09:15<newbie>storrgie: raise a support ticket i gather
09:15<hawk>ericoc: Just out of curiosity, is "data to be moved" the size of the actual raw images?
09:16<@ericoc>hawk: yeah, the empty space is moved too
09:16-!-_Majes [~Majes@] has joined #linode
09:16<HoopyCat>swaj: it's not even so much "go to court" as it is "be a bad network citizen"... it's like the additional-IP thing. linode isn't going to bend so far backwards for a customer that they fall on their ass, you know?
09:16<newbie>bob2: thanks for article link
09:16<HoopyCat>hawk: zeros are data too :-)
09:16<bob2>newbie: I think it means yes
09:16<@ericoc>you can try to minimize the time required by downsizing your disks before-hand
09:16<swaj>HoopyCat: exactly
09:16<newbie>hoopycat: thanks
09:16<storrgie>Peng: done. thanks.
09:16<HoopyCat>ericoc: ITYM "right-size"
09:16<hawk>ericoc: Ok, and no compression or sparse file behavior... got it
09:16<hawk>HoopyCat: Well, it is, but there could be some special handling
09:17<+linbot>Linodes can be migrated between datacenters by simply opening a ticket. Your disk images will be copied intact, but you will be issued a new IP address. Once a migration is configured, it makes a button on the dashboard available for you to start the migration yourself. Migration typically takes about 1-3 minutes per GB.
09:17<HoopyCat>hawk: it's a bit-for-bit copy, so sparse files won't matter... there might be compression, tho
09:17-!-Majes [~Majes@] has quit [Ping timeout: 480 seconds]
09:17-!-_Majes is now known as Majes
09:18-!-cereal|Away is now known as cereal
09:18<newbie>"Lucid upgrade article" seems to be missing thogh
09:19-!-bbeausej [] has joined #linode
09:19-!-bigjocker [~ngranek@] has joined #linode
09:21<newbie>bob2: glad you know what it might mean :) Might stick with what the article you linked to recommends. When stating you add "dev /dev devtmpfs rw 0 0" to /etc/fstab, I assume it means at the end of the file?
09:21<newbie>swaj: ahh..thanks
09:21<bob2>newbie: it means click the box in the web thing that mounts it
09:21<bob2>i think
09:24<@ericoc>yeah, you can either adjust /etc/fstab manually, or enable automount devtmpfs in config profile
09:25<newbie>bob2, ericoc: thought so..but thought it might be better to do it exactly as stated in the add "dev /dev devtmpfs rw 0 0" to THE END of /etc/fstab?
09:27<@ericoc>anywhere in fstab is fine, end is probably more convenient
09:27-!-sean_w [] has left #linode [Konversation terminated!]
09:29<newbie>ericoc: thanks..just wanted to be sure there wasn't something finicky to allow for here.
09:29<newbie>swaj: article is extremely good..actually wished I saw this at the start of this whole process..
09:31-!-Cromulent [] has quit [Ping timeout: 480 seconds]
09:32<swaj>linode library ftw
09:32-!-xt3mp0r [~xt3mp0r@] has joined #linode
09:32-!-copperx [] has joined #linode
09:33-!-jake [] has quit [Server closed connection]
09:33-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
09:33-!-jake [] has joined #linode
09:34-!-vynsynt [] has joined #linode
09:34-!-vynsynt [] has left #linode []
09:34-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
09:34-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
09:35-!-xt3mp0r__ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
09:35-!-xt3mp0r [~xt3mp0r@] has quit [Read error: Connection reset by peer]
09:36-!-redgore [~redgore@] has joined #linode
09:37-!-Rescue9 [] has joined #linode
09:38-!-FooMunki [] has joined #linode
09:39-!-xt3mp0r [~xt3mp0r@] has joined #linode
09:40-!-jayvee [] has joined #linode
09:40-!-copperx [] has quit [Quit: Leaving.]
09:41-!-dajhorn [] has joined #linode
09:42*Daevien facepalms. dude bringing in his comptuer for me to work on. "hotmail" called him to say he had a problem and needed to run software to clean his computer. had his identity stolen, cc used, etc. sad that people actually fall for that
09:42-!-storrgie [] has quit [Quit: CGI:IRC]
09:43*Kuboing FACEPALMS
09:43<Kuboing>they're better off giving everything to us
09:43*czr palmfaces
09:43-!-FooMunki [] has quit [Quit: FooMunki]
09:43<Kuboing>why are people so irresponsible on the internet
09:43<Kuboing>if it's too good to be true; IT FUCKING IS
09:43<czr>because they're also irresponsible in IRL
09:43<Daevien>czr wins that debate :p
09:44<Kuboing>czr; I dont' see them leaving their doors open and putting up signs "ALL MY IMPORTANT STUFF HERE"
09:44<czr>Kuboing, no, you don't see them doing it the first time.
09:44-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
09:44<czr>I'm sure they've learned that lesson by now :-).
09:44-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
09:44<czr>greed + lack of skill = boom.
09:45<Kuboing>czr: so why is everyone still re-using weak passwords across multiple sites after the sony leak, and the lulsec debacle?
09:45-!-FooMunki [] has joined #linode
09:45<czr>you assume that people read about those.
09:45<czr>or care even.
09:45<Kuboing>it has been on the news, ffs
09:46<Rescue9>Kuboing: my personal favorite is this one though:
09:46<Kuboing>CNN, NBC, colbert report, the daily show, etc
09:46<Daevien>jed = lulsec
09:46<Kuboing>Rescue9: I dont' get it
09:47<Rescue9>I dont' see them leaving their doors open and putting up signs "ALL MY IMPORTANT STUFF HERE"......
09:47-!-FooMunki [] has quit [Read error: Connection reset by peer]
09:47-!-FooMunki [] has joined #linode
09:47<czr>Kuboing, you also assume that people watch the news.
09:47<Rescue9>I'd rather use the second sign
09:47-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 481 seconds]
09:47<Kuboing>czr; it has also been on fox
09:47<czr>but yes, I have no idea what makes people stupid and so trusty.
09:47-!-FooMunki [] has left #linode []
09:48<Kuboing>I only occasionally re-use my passwords
09:48<Kuboing>but my passwords are a 32 character MINIMUM
09:48<czr>is it "mypasswordis32characterminimum"?
09:49*Kuboing changes his .gov mail passwords
09:49<HoopyCat>just knowing that something happened from a 20-second news blurb doesn't provide enough context to understand the impact and significance of the event
09:50<Kuboing>HoopyCat: 20 second news blurb? it has been on the TV repeatedly, and for much longer than 20 seconds
09:50<avenj>20 seconds is a long time to pay attention
09:50-!-FooMunki [] has joined #linode
09:51<HoopyCat>Kuboing: people who do business with sony probably use their televisions and might know better than me, indeed
09:51-!-tbye [] has joined #linode
09:51-!-FooMunki [] has left #linode []
09:51<tbye>Can I create a new linode based off the snapshot backup of an existing linode?
09:51<Peng>tbye: Yes.
09:52<tbye>Peng: Could you point me in the right direction in the docs?
09:52<Kuboing>oh-ho! AT&T uses pirated software?
09:52<Peng>tbye: No idea. I just remember that I did it once.
09:52<tbye>Peng: Thanks.
09:52<Peng>tbye: Hold on.
09:52<HoopyCat>tbye: when you click to restore a backup, it'll give you a choice of all the linodes on your account in the same datacenter
09:53<Peng>It also has an "Add a Linode" link.
09:53-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
09:53-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
09:54<tbye>Peng: HoopyCat: Thanks, Gang! Restoring to new linode now.
09:55<newbie>hmmm..going from Ubuntu 9.10 to 10.04 LTS...getting "WARNING: Failed to read mirror file". Thought it was the fact that I had "" in place of /etc/apt/sources.list
09:55<newbie>hmmm..going from Ubuntu 9.10 to 10.04 LTS...getting "WARNING: Failed to read mirror file". Thought it was the fact that I had "" in place of "" & "" in /etc/apt/sources.list
09:55<newbie>changed these back but still getting this warning. thoughts?
09:56<newbie>(does the warning even matter?)
09:57<HoopyCat>hmm, what command is giving that error?
09:58-!-lunks [~lunks@] has quit [Remote host closed the connection]
09:59-!-lunks [] has joined #linode
09:59-!-v0lksman [] has quit [Remote host closed the connection]
09:59<newbie>hoopycat: thanks for reply. "sudo do-release-upgrade --proposed" - paused at asking me if I want to upgrade ie: kept going until asking me if I want to upgrade, hence my question if the warning is one that matters
10:00<newbie>again..I'm going from Karmic to Lucid
10:01<HoopyCat>hmm, i'd switch back to then apt-get update then try the do-release-upgrade again... that said, the warning sounds familiar, but i haven't done an upgrade in over a year
10:01<HoopyCat>the googles might know
10:01-!-Dreamer3 [] has joined #linode
10:04-!-tbye [] has quit [Quit: CGI:IRC (Ping timeout)]
10:04<HoopyCat>i wonder if it is trying to read a file containing a list of mirrors
10:04<newbie>hoopycat: hmmm...tried "sudo apt-get update && sudo apt-get upgrade" then "sudo do-release-upgrade --proposed" - still getting warning
10:04<HoopyCat>newbie: well, it'll probably either work fine or won't work at all
10:05<HoopyCat>it either knows what packages are in the next version, or it doesn't
10:05<newbie>hoppycat: I've checked my /etc/apt/sources.list - I did manually change it back from "" address I needed to go from 9.04 to 9.10
10:05-!-Rescue9 [] has left #linode [Leaving.]
10:06<newbie>hoopycat: fair least I'll know if it breaks..nothing worse than something quietly failing in the background. thanks for thoughts. here goes..
10:06<HoopyCat>oh i dunno, any sort of failure is better than not failing when it should have
10:06-!-Pupeno_ [] has quit [Quit: Pupeno_]
10:06<HoopyCat>try: somethingdangerous()
10:06<HoopyCat>except: pass
10:08<newbie>hoopycat: danger is upgrading a linode..that's dangerous for the linode :)
10:08-!-pigdude [] has joined #linode
10:08-!-Pupeno_ [] has joined #linode
10:08-!-JSharp [] has joined #linode
10:08<HoopyCat>i have a couple systems that gotta do the 9.10->10.04 jump soon... rest assured i'm procrastinating it as much as i can :-)
10:08-!-Pupeno_ [] has quit []
10:09-!-xt3mp0r [~xt3mp0r@] has joined #linode
10:09-!-tiny [] has joined #linode
10:10<newbie>hoopycat: I've had too many experiences of things breaking with an "upgrade" so I know the feeling.
10:10-!-Pupeno_ [] has joined #linode
10:11-!-iurisilvio [~iurisilvi@] has joined #linode
10:11<HoopyCat>newbie: everything "new" is deployed/maintained using chef, but damn if the old stuff doesn't still pay the bills
10:11-!-copperx [] has joined #linode
10:11<newbie>hoopycat: chef
10:12<HoopyCat>one of the frustrating things about business in the internet age... happened at $UsedToWorkThere too. awesome fancypants new VoIP service with lots of new customers, but you look at the financials, and frickin' modem dialup is still where the money is...
10:12-!-lrojas [] has joined #linode
10:12<lrojas>hi all
10:12*newbie watcing linode upgrade away..
10:13<newbie>lrojas: hey
10:13<AviMarcus>modem dialup? that still exists?
10:13<lrojas>i am trying to set my host name ( my linode is debian 6 ), and i have followed the instructions 3 times in the guide...
10:13*newbie prefers to be in the if it ain't broke don't fix it camp
10:14-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
10:14<newbie>unfortunately..people like to work out ways of breaking things
10:14<newbie>avimarcus: museum :)
10:15<HoopyCat>newbie: ... "configuration as code" is the best way i can explain it. you write code that configures your systems. :-)
10:15<HoopyCat>lrojas: does it keep reverting to the default? i think the DHCP client is configured to reset it, by default...
10:16<HoopyCat>AviMarcus: it did when i left a number of years back :-) i still have the local modem number memorized tho...
10:16<HoopyCat>AviMarcus: still alive and well
10:16-!-linville [] has joined #linode
10:18<auraka>hey hoopy
10:18<newbie>hoopycat: i see. well..why do it manually when you can automate. nice. stack more configuration files that are different in the 9.10 to 10.04 upgrade vs 9.04 to 9.10
10:19<HoopyCat>AviMarcus: we used to actually have max4048s to handle this pool, and we did actually replace them with TNTs to go V.92, but we ported all of our modem numbers to a wholesaler and scrapped the banks before i left
10:19<AviMarcus>I don't know what any of that means. Other than port to wholesaler.
10:20<HoopyCat>good morning, auraka
10:21-!-s_t [] has joined #linode
10:23<HoopyCat>newbie: yeah, we kinda have to automate. i'm responsible for, uhh, 27 servers? wow, damn
10:23<HoopyCat>when the hell did it get above 20
10:23<s_t>can i have a coupon code
10:24<HoopyCat>s_t: they are very rare, and tend to be single-use when they do exist
10:24<auraka>s_t: this isn't dairy queen
10:24<Peng>Ack I want a Blizzard now.
10:25-!-Kebn [] has joined #linode
10:25*Peng shakes fist at auraka
10:26<HoopyCat>newbie: when you can ssh into each machine to do updates manually, you don't have many servers. when you have a fabfile that sshs into all of your machines to do updates, you've got more than a handful of machines. when your fabfile uses libcloud to automatically discover all of the machines you have, you're starting to get big
10:26-!-comhack [] has quit [Quit: Lost terminal]
10:26<s_t>@auraka you have opened my eyes ...
10:26-!-copperx [] has quit [Quit: Leaving.]
10:27-!-Kebn [] has quit []
10:27<s_t>@Hoppycare thanks signing up anyway
10:27<auraka>go forth and conquer the world now
10:27<lrojas>the changes take effect for a few minutes, then revert back to default, i have set the /etc/default/dhcpcd file to comment out the SET_HOSTNAME='yes'
10:27<HoopyCat>s_t: if there was someone who referred you to linode, you can use their referral code and they'll get a $20 credit if you stick around for >90 days :-)
10:27<Peng>HoopyCat: Your 27 servers are all Linodes, right? :D
10:27<lrojas>HoopyCat: yes, they keep reverting during the night to the default naming
10:28<s_t>Linode turns 8; Disk space +25%
10:28<auraka>and hopefully in different data centers
10:28<JediMaster>heh, it's 29C outside and it's hailing =)
10:28<HoopyCat>Peng: alas, no. not my decision. i did, however, architect everything so that we could be on linode within a couple hours :-)
10:28<AviMarcus>HoopyCat, so the upgrade is a good time to do so? ;)
10:28<HoopyCat>Peng: 2 of them are linodes and 1 is a physical box
10:29<auraka>how do you get 27 from 2 linodes and a physical box?
10:29-!-verb [~blarg@2600:3c01::f03c:91ff:fe93:da] has joined #linode
10:29<lrojas>is anybody here an official representative from linode?
10:29<Peng>...The other 24 are different things?
10:29<Peng>!ops lrojas
10:29<+linbot>Users with ops are employees of Linode, and know what they're talking about. The rest of us are the ever-so-helpful(?) community. Official Linode contact information:
10:29<Peng>lrojas: Why do you ask?
10:29<HoopyCat>lrojas: hmm... is the reverse DNS set? i wonder if that might help it.
10:30<lrojas>i am not hosting my dns with linode
10:30-!-userme [] has joined #linode
10:30<Peng>lrojas: HoopyCat didn't say you were.
10:30<lrojas>Peng: i have set my hostname severla times now, as per the instructions on the guide
10:30<HoopyCat>lrojas: that's ok; if there's an A record pointing at the linode's IP, the reverse DNS setting on the remote access tab will pick it up
10:31<s_t>no one referred me here... linode's reputation on WHT
10:31<auraka>lrojas: did you set it in the /etc/hostname file as well as putting it in /etc/hosts for good measure
10:31<lrojas>auraka: yup
10:31<auraka>instead of commenting out SET_HOSTNAME why not change it to the default may be yes
10:31<newbie>heeellloooo...Lucid :)
10:31<auraka>if you just comment it out it will go to default
10:31<HoopyCat>s_t: dang crowdsourced referrals
10:31<newbie>10.04.2 LTS goodness here we are
10:31<lrojas>auraka: good call, i will try that.. is just that.. this is just weird...
10:32-!-alex_m [] has joined #linode
10:32<newbie>my thanks to all who helped me wrestle this puppy home
10:32<pronto> now _that_ is a LOT of states... o_o
10:32-!-febits [] has joined #linode
10:33<HoopyCat>newbie: yay!
10:34<lrojas>well i changed it again, lest hope it stays like that now
10:35<newbie>hoopycat: seriously..THANKS...wish the other guys who helped out were here to thank but there it is.
10:36<auraka>newbie: I accept unearned thanks
10:36-!-s_t [] has quit [Quit: CGI:IRC (Ping timeout)]
10:36<HoopyCat>newbie: next time you're in rochester, i'll gladly accept beer on behalf of everyone else
10:37<HoopyCat>(afk, maintenance window)
10:37<newbie>hoopycat: it won't be just 1 beer :) which rochester exactly? NY?
10:38<Peng>You have to buy HoopyCat beer in every Rochester!
10:38<newbie>auraka: plenty to go around :)
10:38<newbie>peng: haha
10:39<HoopyCat>newbie: yup. target and give me a few hours' warning ;-)
10:39<newbie>bob2: hopefully you'll be around sometime to see this but thanks to you too
10:39-!-xt3mp0r [~xt3mp0r@] has quit [Read error: Connection reset by peer]
10:39-!-xt3mp0r_ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
10:39<newbie>hoopycat: nice..but it'll be more than a few hours given I'm not even in that continent
10:40-!-freshmilk [~work@] has quit [Quit: freshmilk]
10:41<Peng>What about Rochester, Victoria? Or you could meet on USS Rochester, wherever it is.
10:41<+linbot>New news from wiki: User:Merriam <>
10:41<Peng>Ah, damn, scrapped.
10:42<HoopyCat>newbie: well, it's ~$200 round-trip from KJFK to KROC via JetBlue, so if you do it, i'll buy you a pint to cover the cost :-)
10:43<Peng>That's quite a pint.
10:43<HoopyCat>Peng: it does look like quite a pint
10:43-!-xt3mp0r [~xt3mp0r@] has joined #linode
10:44-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
10:44<HoopyCat>(seriously, afk)
10:48-!-dajhorn [] has quit [Remote host closed the connection]
10:48-!-dajhorn [] has joined #linode
10:50<newbie>hoopycat: keep in never know if I'm ever at that end of the woods :)
10:52-!-JediMaster [] has quit [Ping timeout: 480 seconds]
10:52-!-JediMaster [] has joined #linode
10:54-!-streety [] has left #linode []
10:54-!-vynsynt [] has joined #linode
10:54-!-vynsynt [] has left #linode []
10:55-!-tjfontaine [] has joined #linode
10:56-!-ktabic [] has quit [Quit: I'm a professionally trainined computer scientist. That is to say, I am poorly educated]
10:57-!-imroot702 [] has quit [Quit: Leaving]
10:58-!-v0lksman [] has joined #linode
10:59-!-streety [] has joined #linode
10:59-!-aussiekid [] has quit [Ping timeout: 480 seconds]
11:02-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
11:03-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
11:03-!-newbie [] has quit [Quit: CGI:IRC (Ping timeout)]
11:04-!-karstensrage [] has quit [Quit: Leaving]
11:07-!-xt3mp0r [~xt3mp0r@] has joined #linode
11:11-!-shirro [] has quit [Quit: shirro]
11:13-!-tempesta [] has joined #linode
11:15-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
11:16<JediMaster>mmmm, rsyncing 400,000 files to one server, and they magically appear on the FS of the second server instantly =)
11:17*JediMaster is liking GFS2 so far
11:18<Peng>You liked the first 10 days? :P
11:18<JediMaster>that wasn't GFS
11:19<JediMaster>that was kernel compiling, linode configuring, DRBD compiling and mucking aroudn trying to get multicasting working
11:19<JediMaster>which were all prerequisites to OCFS2 which sucked up an entire day of work, before I tried GFS2 which took all of 5 min
11:20<@jed>at least you chose the right messaging layer (corosync) first, without having to migrate to the right one
11:20<@jed><-- did heartbeat -> corosync
11:20<JediMaster>I'm glad I had the foresight to compile the kernel with GFS2 as well as OCFS2 =)
11:20<JediMaster>just setting up mysql multi-master replication now
11:21<tjfontaine>ha ha mysql
11:21<JediMaster>is there any way around having to off-set the autoincrement ids on each master so they don't conflict when one goes offline?
11:21<@jed>JediMaster: do you like torture or something?
11:21*jed shivers
11:22<JediMaster>jed: got any better ideas for multiple mysql servers?
11:22<JediMaster>can't use mysql cluster as the application (magento) doesn't support it
11:23-!-xt3mp0r [~xt3mp0r@] has joined #linode
11:23<JediMaster>I'm hoping to convince the developers, some of whom I've actually met (we're a partner) to add mysql cluster support in magento 2
11:24<pharaun>jed: i think we have someone here who likes torture :3
11:24<JediMaster>I'm mostly a php developer, but I'm one of the odd ones that likes screwing around with odd server setups and sysadmin weirdness
11:27<HoopyCat>JediMaster: they have to be offset, alas, since multi-master replication is just a loop of master-slave replications...
11:27<HoopyCat>JediMaster: does mysql cluster need application support? afaik, it just presents a "normal" mysql server interface to the application...
11:27<JediMaster>yeah, I guess what I really need is mysql cluster that works on innodb
11:28<tjfontaine>ha ha ha
11:28-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
11:28-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
11:28<tjfontaine>there's so much troll bait in this conversation
11:31-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
11:32-!-bbeausej [] has quit [Read error: Operation timed out]
11:32-!-iurisilvio [~iurisilvi@] has quit [Ping timeout: 480 seconds]
11:36<JediMaster>HoopyCat: afaik it's a seperate table engine, so isn't compatible with innodb tables
11:39-!-bbeausej [] has joined #linode
11:39-!-xt3mp0r_ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
11:39-!-xt3mp0r__ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
11:40-!-walterheck_ [~walterhec@] has quit [Quit: Computer has gone to sleep]
11:40-!-lrojas [] has quit [Remote host closed the connection]
11:43-!-walterheck [~walterhec@] has joined #linode
11:44-!-xt3mp0r [~xt3mp0r@] has joined #linode
11:44-!-atula [] has quit [Read error: Connection reset by peer]
11:44-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
11:45-!-atula [] has joined #linode
11:46-!-foreverwondering [] has joined #linode
11:46-!-foreverwondering [] has left #linode []
11:48<+linbot>New news from forums: Question about Linode Backup Service - how on-site is it? in General Discussion <>
11:50<h3lix>i just allocated a new linode 512, and it says I have 20GB of transfer per month.
11:50<h3lix>is that because it's prorated to the end of the month or something?
11:51-!-tiny [] has quit [Ping timeout: 480 seconds]
11:51<mdcollins>Yeah, it's prorated
11:51<h3lix>cool. thanks
11:51<pronto>because all the pro's use it!
11:51-!-kraz [] has quit [Remote host closed the connection]
11:51*pronto hides
11:51<h3lix>you'd better hide.
11:51-!-kraz [] has joined #linode
11:52<swaj>so I bought an Atom D510-based dedicated box for windows server 2008. Pretty pleased so far :P
11:52<swaj>they set up a RAID 1 mirror for me
11:52<pronto>oh nice, that ssimliar to what im u sing for pfsense
11:52<swaj>the provider
11:52-!-foreverwondering [] has joined #linode
11:53-!-foreverwondering [] has left #linode []
11:53<swaj>I wanted to be protected against hard drive failure -- the box came with a 500 GB sata drive, and they only charged me an extra $5/month to put in a second one and set up RAID 1 for me
11:53<mdcollins>I have a similar for my home xen server.
11:53-!-stafamus [~stafamus@] has quit [Ping timeout: 480 seconds]
11:54<swaj>Atom D-510, dual 500 GB drives, and 4 GB RAM
11:54<swaj>not a bad little web server :P
11:54-!-chmac [] has joined #linode
11:54<swaj>it's slow installing updates, but plenty fast for serving up some ASP.NET MVC stuff
11:55-!-Jippi [~jippignu@] has quit [Quit: Jippi]
11:55<mdcollins>That's very similar to my server, but raid1 1.5tb drives.
11:55<swaj>yeah I honestly don't need all that space. It's just what the server came with. It's only meant to host some .NET stuff for me.
11:56<hawk>mdcollins: Xen and Atom isn't something that one exactly think of in the same sentence... But it works out ok? I guess the load isn't that bad in the first place...
11:56<swaj>I thought the same :P
11:56<swaj>Atom doesn't have VT-x
11:56<swaj>so the guests would have to be modified
11:56<mdcollins>Oh, mine's for home storage.
11:57-!-tunabananas [] has joined #linode
11:57<hawk>mdcollins: Ah, I thought you said Xen
11:57-!-dstites [~dstites@] has joined #linode
11:57<mdcollins>yeah, that too
11:57<swaj>ah, there are some Atom procs that do Xen
11:57<swaj>the Z5XX series and the EXXX series
11:57<swaj>err I mean VT-x
11:58-!-sivy [] has joined #linode
11:58<hawk>swaj: Ah, ok
11:58<swaj>though they're single cor
11:58<swaj>single core*
11:58<mdcollins>I can only do paravirt, but its for my personal use.
11:59<mdcollins>Mine is dual core.
11:59<swaj>it appears the dual core Atoms don't support VT-x at all. Like the D510 I've got. Dual Core, HT so my OS sees 4 "procs"
12:00<hawk>swaj: I have one machine with an older dual core Atom (330?)... But it's kind of underwhelming overall, which I guess is to be expected for something that was built specifically to be cheap and low power.
12:00<swaj>yeah the 330 is dual core
12:00-!-chmac [] has quit [Remote host closed the connection]
12:00<swaj>Atoms are not power houses for sure. But for a basic entry-level dedicated server used for simple web hosting, it works out well.
12:01<hawk>I have a feeling the chipset is really crappy though, that it's not really the slowish cpu that bogs it down the most but things like disk io, etc
12:01<swaj>the hardware on my little server is SuperMicro, actually
12:01<swaj>pretty good stuff
12:01<tunabananas>I am looking to install phpmyadmin to restore a wordpress sql backup, using Ubuntu 10.04 on MacOS.. can anyone point me to best documentation?
12:02-!-chmac [] has joined #linode
12:02<tunabananas>or perhaps there is an alternative to phpmyadmin..
12:02<swaj>looks like I'm on an ICH9 north bridge
12:02<swaj>tunabananas: mysql workbench over an SSH tunnel.
12:03<hawk>swaj: That sounds like a better base :)
12:03<swaj>aye :P
12:03<tunabananas>w0rd~ so i would start here?
12:04<swaj>MySQL workbench will create the tunnel on your end, for you
12:04<swaj>when you add your server, just set it up as "Standard TCP/IP over SSH"
12:04<mdcollins>I was more focused on electricity usage and noise, seeing as it's in my living room and on 24/7. But it's an awesome little machine.
12:04<swaj>and set the remote host to your server's IP... and then when it asks for the sql server's ip relative to the remote host, just use (localhost)
12:05-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
12:06-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
12:06<tunabananas>sweet, SUPERGRACIAS swaj!
12:06<swaj>this way you don't ever need that horrible gaping security hole called phpMyAdmin :P
12:07<tunabananas>yes~ precisely why i didn't set this up before... am remembering admonishments from my softwaredev buddy...
12:07<mdcollins>I use http auth..
12:07-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
12:07-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
12:08<swaj>yeah BASIC auth goes a long way, but I still would rather not even have PMA. the MySQL workbench is really a great tool
12:08<swaj>(though I'm debatig switching all my stuff over to pgsql)
12:10<dstites>is this an OK place to ask technical questions? i have followed some of the articles on the library re: apache setup but my configuration is a bit more complicated than what they show and i am having some problems getting my vhost config to serve the proper directory.
12:10-!-seanh-ansca [] has joined #linode
12:10<+linbot>If you have a question, please just ask it. Don't look for topic experts. Don't ask to ask. Don't PM! Don't ask if people are awake, or in the mood to help. Just ask the question straight out.
12:12-!-imroot702 [] has joined #linode
12:12<dstites>alright then, i have several domains and several rails apps. domains a and b should go to rails app 1, and domains c and d should go to rails app 2. it appears that my dns is configured properly because in some instances, i do serve the proper site. in other cases, such as when i add a www onto the front of the domain, it serves the improper site
12:13<dstites>unfortuantely, i can't really check apache error logs because technically there is no error - it is just choosing the wrong vhost to serve it from
12:13<swaj>do you have ServerAlias directives set in the <VirtualHost>?
12:13<dstites>yes. i list all the domains first without the www prepended to the domain name, and then i list
12:13<dstites>as a list
12:14<swaj>they each need their own line
12:14<dstites>interesting. i have seen some of the apache configs where they are all listed on the same line, even doing wild cardmatching like *
12:14<swaj>my apache knowledge is a bit old, though
12:15-!-kenichi [] has joined #linode
12:15<swaj>perhaps that's possible now
12:15<dstites>would it be correct to list the domain without the www subdomain first?
12:15<swaj>(I use nginx) :P
12:15<dstites>or would it just be easier to resolve by getting an additional IP and doing an IP based vhost instead of a name based vhost
12:16<swaj>looks like you can have multiples, although the wildcard appears to be by itself
12:16<swaj>your name-based method should work. As long as it's not SSL :P
12:16<swaj>for https you need unique IPs for each site
12:16<tjfontaine>or SNI
12:16<dstites>no, it's not. its all regular http
12:17<swaj>so yeah, there's gotta be something off with the serveralias lines
12:17-!-freshmilk [~work@] has joined #linode
12:17<dstites>one thing that i did note is that when i did an apachectl -S, it did list the default as the one it serves when i tag www onto the domain
12:18<swaj>if it can't find a matching alias, then it uses the default for all requests
12:18<dstites>ah alright - would mod_rewrite possible affect this?
12:19-!-ktabic [] has joined #linode
12:20<dstites>or would splitting up the vhost files into multiple ones affect it?
12:20<dstites>the logic behind that is i could enable or disable sites individually, instead of all sites at once
12:20-!-_Majes [~Majes@] has joined #linode
12:21-!-epochwolf [] has joined #linode
12:21-!-epochwolf [] has quit []
12:22<dstites>swaj, in that link you provided, i don't have a UseCanonicalName if that might affect it
12:22-!-Majes [~Majes@] has quit [Ping timeout: 480 seconds]
12:22-!-_Majes is now known as Majes
12:25-!-hipsterslapfight [] has quit [Quit: hipsterslapfight]
12:26-!-Boohemian [] has quit [Quit: Computer has gone to sleep.]
12:26-!-comhack [] has joined #linode
12:28-!-kenichi [] has quit [Remote host closed the connection]
12:35-!-copperx [~Adium@] has joined #linode
12:36-!-JediMaster [] has quit [Ping timeout: 480 seconds]
12:36-!-Kebn [] has joined #linode
12:36-!-JediMaster [] has joined #linode
12:37-!-CallMeD [~CallMeD@] has joined #linode
12:44-!-freshmilk [~work@] has quit [Quit: freshmilk]
12:45-!-djg320 [] has joined #linode
12:45<digilink>server -m sbrown-hak5:k1lnxar0 -i digilink digilink digilink
12:45-!-CallMeD [~CallMeD@] has left #linode [Leaving]
12:45<hawk>Missing slash perhaps?
12:45-!-Cromulent [] has joined #linode
12:46-!-digilink [~digilink@2600:3c03::f03c:91ff:fe93:f26c] has quit [Quit: "Bite my shiny metal ass!" --Bender]
12:46<pronto>and that's why you always use the status window to put in commands like that...
12:47<dominikh>I am going to bet my ass on that he didn't change the login details
12:47*pronto slaps mikegrb
12:47*pronto tosses kittens at SpaceHobo
12:47-!-squircle [] has joined #linode
12:48-!-dstites [~dstites@] has quit [Quit: dstites]
12:48-!-JediMaster [] has quit [Ping timeout: 480 seconds]
12:48-!-JediMaster [] has joined #linode
12:48*squircle goes to change his nickserv password
12:50-!-vraa [] has quit [Ping timeout: 480 seconds]
12:50<copperx>I'm so using that password for my stuff
12:51-!-nisstyre [~nisstyre@] has joined #linode
12:51<squircle>at least I don't use that password for anything else
12:51<copperx>nice try, squircle
12:51<squircle>and, really, it's IRC... not the end of the world (doesn't mean I'm not embarrassed, but...)
12:51<hawk>squircle: Hey, how come you have my password?!
12:52<squircle>hawk: I don't anymore! :)
12:53-!-nicinabox [] has joined #linode
12:53-!-nisstyre [~nisstyre@] has quit []
12:53-!-djg320-server [] has quit [Ping timeout: 480 seconds]
12:53-!-nisstyre [~nisstyre@] has joined #linode
12:54-!-febits [] has quit [Read error: Connection reset by peer]
12:55-!-febits [] has joined #linode
12:56<+linbot>New news from forums: looking for someone who knows the API in General Discussion <>
12:57-!-bbeausej [] has quit [Quit: Leaving.]
12:59<DrJ>you know caker ... speaking of API ... with the API you have for the dns available and no limits on it ... it would be really easy for someone to start a free dns service using it ;)
12:59-!-bbeausej [] has joined #linode
13:00<hawk>Except they may suddenly lose the whole foundation for their service if their use violates the terms
13:00<DrJ>not so sure it does
13:01<tjfontaine>or if jed gets cranky
13:01*DrJ could probably get something like that online and working in a couple hours
13:03-!-Cromulent [] has quit [Remote host closed the connection]
13:04-!-copperx [~Adium@] has left #linode []
13:04<DrJ>I'm sure if someone tried the hammer would come down though
13:06-!-bonjurkes [] has joined #linode
13:07<bonjurkes>hey everyone, i decided to switch to nginx instead of apache and followed this guide : i am on ubuntu 11 tho. At /etc/init.d/php-fastcgi file configuration it gives error about bin/bash. what am i doing wrong?
13:09-!-capsl0ck [] has joined #linode
13:09<retro|blah>"error about bin/bash" What error?
13:10<hawk>DrJ: And caker will strike down upon thee with great vengeance and furious anger those who would attempt to misuse or abuse the dns service...
13:11<bonjurkes>bad interpreter : /bin/bash no such file or directory
13:11<bonjurkes>when i try to start phpfastcgi
13:11-!-hfb [] has joined #linode
13:12<hawk>bonjurkes: I suppose it requires bash... and you don't seem to have it (yet?)
13:12<retro|blah>What shell are you using then!?
13:12<hawk>Not everyone like bash... But in this case I suppose he'll need it for some non-interactive things
13:13<bonjurkes>i am using zsh perhaps thats the reason?
13:13<hawk>bonjurkes: Not because you use zsh, but because bash is not available on the system
13:13<hawk>bonjurkes: And apparently the script is written for bash
13:13-!-JSharp [] has quit [Quit: Leaving]
13:14<bonjurkes>its just what written on linode's library, not sure is it ok if i use different one
13:15-!-hipsterslapfight [] has joined #linode
13:15-!-hipsterslapfight is now known as spkitty
13:15<hawk>bonjurkes: If there's nothing bash specific in the script, change the shebang to say /bin/sh instead
13:16<bonjurkes>apt-get says bash is already latest version. hawk then it says same thing for bin/sh
13:16<hawk>Does /bin/bash and/or /bin/sh exist?
13:16<bonjurkes>damn sudo
13:17<JediMaster>you about jed?
13:17<bonjurkes>i have to type sudo then the command :P
13:18<JediMaster>jed: I've got a nodebalancer setup, and I changed it to source ip balancing, then switched it back to round-robin, but it's stuck doing source-ip based balancing even though the interface says otherwise
13:19-!-tempesta [] has quit [Ping timeout: 480 seconds]
13:21-!-tempesta [] has joined #linode
13:23-!-Jippi [] has joined #linode
13:25-!-djg320 [] has quit [Ping timeout: 480 seconds]
13:31<@jed>JediMaster: how do you know?
13:32<@jed>er, you're on stickiness table
13:32<@jed>so the decision about where to land gets made round-robin, then the balancer remembers that source IP and gives it the same backend every time
13:34-!-JediMaster [] has quit [Ping timeout: 480 seconds]
13:34-!-JediMaster [] has joined #linode
13:35-!-byronb [] has joined #linode
13:35<@jed>JediMaster: I show you changed the config at 7:48:06 Linode time, and the balancer was reconfigured 4h28m ago
13:35<hawk>JediMaster: A disturbance in the force?
13:35<@jed>it is now 11:35:21 Linode time
13:35<@jed>er, 13
13:35*jed walks away
13:35<tjfontaine>put the bottle down
13:36<@jed>JediMaster: the stickiness is what's causing the behavior you're seeing
13:38-!-sm [~sm@] has joined #linode
13:40<navi>Linode has a timezone?
13:40<tjfontaine>it's own
13:40<navi>So... the london servers run on Linode Time, or GMT/BST?
13:40<@jed>root@london1:~# date
13:40<@jed>Tue Jun 28 13:40:40 EDT 2011
13:40<tjfontaine>servers run in LinodeTim which is unlike any other time in the universe
13:40<navi>That's STOPPID.
13:41<@jed>it is?
13:41<tjfontaine>navi: omig did they hurt your feelings?
13:41-!-tempesta [] has quit [Ping timeout: 480 seconds]
13:41<navi>Well, I guess it's not that bad, I mean, when the status page shows an issue, it does show UTC sometimes
13:42<tjfontaine>wah wah wah, tough shit
13:42<tjfontaine>addition and subtraction, it's hard
13:42<navi>tjfontaine: It is kinda hard when you're told "EDT" and not what to add or subtract.
13:42<@ericoc>ntp <3
13:43<tjfontaine>date will tell you, so will google, and wolfram and and and and and you're helpless
13:43<@jed>root@london1:~# TZ=UTC date
13:43<@jed>Tue Jun 28 17:43:08 UTC 2011
13:43<@ericoc>your linode will only reference the host's time if you're using an older kernel
13:43<@jed>root@london1:~# date -u
13:43<@jed>Tue Jun 28 17:43:41 UTC 2011
13:43<@jed>I am skinning many cats!
13:43<hawk>poor cats
13:44<navi>jed: I thought we had this conversation about a week ago, that's illegal
13:44<@jed>what else will I put in my head in TF2
13:44<@jed>on, rather
13:44<navi>Jed puts cats in his head!
13:44<@jed>my brain is entirely cats
13:44<hawk>skinned cats
13:44-!-pigdude [] has quit [Ping timeout: 480 seconds]
13:45<navi>Cat, I'm a kitty cat. And I dance dance dance and I dance dance dance...
13:45-!-tempesta [] has joined #linode
13:47-!-bcum1 [~k@] has joined #linode
13:48<JshWright>navi: not illegal if the cat is already dead from some other cause
13:48<JshWright>though in many places, you need a taxidermy license
13:48-!-pigdude [] has joined #linode
13:49-!-sm [~sm@] has left #linode []
13:49-!-vraa [] has joined #linode
13:49<navi>I wonder what the laws actually are about what you can and can't do to a dead cat
13:50<navi>I mean, I guess you can skin and stuff it
13:50<HoopyCat>you may not, under any circumstances, pipe a cat
13:50<tjfontaine>you can make kitten mittens
13:50<navi>But I'm sure the authorities would still not like it if you, for example, ran over it with a steamroller
13:50<bonjurkes>still following : tutorial, spawn-fcgi says cant find username www-data, but there is www-data username !
13:50<navi>"In case it was still capable of feeling" or something
13:51*navi has a spare steamroller and is waiting for that to become legal
13:51<hawk>Steam rollers are illegal?
13:51<navi>How about locking it in a cage and having people vote on... nah, that's a stupid idea. Nobody'd ever do that.
13:52<hawk>navi: :>
13:52-!-bcum1 [~k@] has left #linode []
13:52<navi>If that ever happened, I'm sure they'd stop it immediately(!)
13:53<navi>I'd boycott any company that would suppor... Uh... sorry, what was I saying?
13:53<navi>Oh, right, steamrollers.
13:54<hawk>Can we vote for what you should do with your spare steam roller?
13:54<navi>KILL or SAVE the steamroller?
13:54<swaj>bonjurkes: I'd be using php-fpm instead of spawn-fcgi
13:55<hawk>SpaceHobo: Uh?
13:56<navi>Oh shi-
13:56*navi has just caused himself a nosebleed by breathing too hard.
13:58<hawk>If you bleed to death, can we skin you?
13:59<tjfontaine>I want to skin him while he is alive and aware
14:01-!-jxpx777 [~jxpx777@] has joined #linode
14:01*navi returns to see some hurtful comments and begins crying
14:02-!-vraa [] has quit [Read error: Connection reset by peer]
14:02<hawk>tjfontaine: Is that legal? Maybe that only applied to cats...
14:02<+linbot>urmom is *always* relevant.
14:02<+linbot>navi: Yo mommas so stupid, she keeps clicking the link to (835:7/0) [romum]
14:03<deejoe>you know what's weird? That even with all the 'urmom' stuff, this channel is way less blatantly offensive than #othervpsprovider
14:03<tjfontaine>how excellently referential linbot
14:05-!-bbeausej1 [] has joined #linode
14:05-!-bbeausej [] has quit [Read error: Connection reset by peer]
14:08<bonjurkes>swaj thanks !
14:08<hawk>deejoe: Maybe thanks to rather than despite?
14:11-!-loxs [~loxs@] has joined #linode
14:12-!-advion [] has joined #linode
14:12-!-loxs [~loxs@] has quit []
14:13-!-nellboy [] has joined #linode
14:15<nellboy>Hi, I'm currently moving from slicehost to linode, and would really like to just setup my linode from a slicehost backup, I doubt this is possible, but maybe it is?
14:15<nellboy>anyone have any info/suggestions ?
14:16<retro|blah>!library migrate
14:16<+linbot>retro|blah: 1. Migrate a Server to Your Linode - | 2. Migrate Disk Images and Linode Instances - | 3. Linode Beginner's Guide -
14:17<retro|blah>I haven't had to do this personally, but many others have it appears
14:17-!-amar__ [] has joined #linode
14:19-!-Cromulent [] has joined #linode
14:19<amitz>!library migraine
14:19<+linbot>amitz: No results. :(
14:19*amitz returns to lurking ;-)
14:20<nellboy>hmmm, looks about as tricky and problem prone as just setting up the slice and copying the database and files over ...
14:20<nellboy>think i'll just go the setup route
14:20-!-xt3mp0r [~xt3mp0r@] has joined #linode
14:20<nellboy>thanks anyway
14:20-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
14:22-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
14:22-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
14:24-!-Tobarja [~chatzilla@] has joined #linode
14:24-!-nellboy [] has quit [Quit: nellboy]
14:24-!-rns [] has joined #linode
14:26-!-lakin [] has joined #linode
14:27<bonjurkes>do i need to install php-fastcgi to use php fpm ?
14:29-!-TIBS01 [lemfpomwe@] has joined #linode
14:34<hawk>I think it's kind of a replacement... but I'm not really that into these php things
14:35-!-jxpx777 [~jxpx777@] has quit [Quit: Mac is sleeping…]
14:37<bonjurkes>yeah its replacement, just realized. Now i should find why php-fpm pool binds to 9000 port
14:38-!-fcoury [] has joined #linode
14:39<Zr40>what was linbot's weather command?
14:39<@jed>!wx 08205
14:39<+linbot>jed: [metar] 08205: not found
14:39<@jed>!wx KACY
14:39<+linbot>jed: [metar] OBS at KACY: 75.2F/24C, visibility 3 miles, wind 9.21 mph (altimeter: 29.9) [KACY 281824Z 05008KT 3/4SM +RA BR FEW002 BKN041 OVC047 24/23 A2990 RMK AO2 P0007 ]
14:39<Zr40>!wx EHEH
14:40<+linbot>Zr40: [metar] OBS at EHEH: 77.0F/25C, visibility 1600 miles, wind 29.92 mph (altimeter: 29.854812813) [EHEH 281829Z AUTO 24026G44KT 1600 R22/P3000N -TSRA FEW024CB SCT068 BKN074 25/19 Q1011 YLO BLU TEMPO YLO 14008KT 9999 BKN120 TEMPO 24018G38KT 2000 TSRAGR SCT035 BKN045CB]
14:40-!-JediMaster [] has quit [Ping timeout: 480 seconds]
14:40<Zr40>29.92 mph!
14:40<@jed>1600 mile visibility?
14:40<Zr40>yesterday it was at 9999
14:41-!-walterheck [~walterhec@] has quit [Quit: Computer has gone to sleep]
14:42<DrJ>!wx larwill
14:42<+linbot>DrJ: [metar] LARWILL: not found
14:42<DrJ>!wx 46764
14:42<+linbot>DrJ: [metar] 46764: not found
14:42<DrJ>!wx hell
14:42<+linbot>DrJ: [metar] HELL: not found
14:42<@jed>linbot works in PM
14:42<@jed>and you want an ICAO code
14:43<DrJ>not sure what icao is
14:43<DrJ>!wx fwa
14:43<+linbot>DrJ: [metar] FWA: not found
14:43<@jed>!wx KFWA
14:43<+linbot>jed: [metar] OBS at KFWA: 77.0F/25C, visibility 10 miles, wind 18.41 mph (altimeter: 29.97) [KFWA 281754Z 29016G22KT 10SM FEW055 SCT200 25/12 A2997 RMK AO2 SLP144 T02500117 10256 20217 50007 ]
14:44<bonjurkes>i am checking netstat grep 9000 port and it says main.conf using it, what does that mean ?
14:44<fo0bar> <-- Linode anniversaries are like Oprah giveaways
14:44<fo0bar>You get a space increase, you get a space increase, EVERYBODY GETS A SPACE INCREASE!
14:44*DrJ can't wait till the 10th birthday
14:44<DrJ>I have a feeling caker might be feeling extra giveful that day
14:44<tjfontaine>fo0bar: when does it launch its own tv network?
14:45<@mikegrb>mmm cake
14:45<Zr40>DrJ: I have a feeling it involves cake
14:45<Daevien>drj: or caker will screw with everyone and take away space for the 10th :
14:45*DrJ has got to hand it to linode though ... most providers only change one things on their plans
14:45<DrJ>their price
14:45<fo0bar>tjfontaine: I hear negotiations with the Cybersauce Broadcasting Corporation are ongoing
14:45<Zr40>DrJ: up?
14:46<fo0bar>jed: I loved that one -- it was a parody on Conan
14:46*fo0bar prepares to activate native ipv6 on a linode
14:47<fo0bar>what's up with the MAC change, by the way?
14:47*Zr40 has been preparing (read: waiting) for that since the announcement (@ atlanta :P)
14:47<Zr40>fo0bar: MAC change?
14:47<fo0bar>> IPv6 requires a change to your Linode's MAC address and a reboot. After submitting this form, the MAC change will take effect immediately and your Linode will be rebooted. You must check the checkbox below to continue.
14:47<fo0bar>from the manager
14:48<Zr40>fo0bar: ah, I guess the old MAC wasn't unique, but that didn't matter on IPv4-only hosts
14:48<Zr40>fo0bar: still, a MAC change won't affect anything :P
14:48<@jed>fo0bar: new hotness.
14:48-!-HIghoS_ is now known as HIghoS
14:50<fo0bar>Zr40: actually, on debian systems, it'll cause the interface to be renamed in some cases, since debian's udev setup likes to write the mac to its conf (annoying when you need to change nics, but useful if you have multiple nics, so they don't jump around due to other conditions)
14:50<fo0bar>SUBSYSTEM=="net", DRIVERS=="?*", ATTRS{address}=="fe:fd:42:dc:01:f7", NAME="eth0"
14:50<fo0bar>still, quick fix, just a matter of deleting /etc/udev/rules.d/70-persistent-net.rules
14:50<Zr40>fo0bar: good to know
14:52<Zr40>I wonder. If I create a new Linode in a DC with IPv6 but don't choose to enable IPv6, would I need to change the MAC once I enable IPv6?
14:52<fo0bar>Lassie initiated boot - System shutdown - System boot <-- serves me right for shutting it down manually before clicking "OK"
14:52<@caker>for now, anyway.
14:53<kenyon>oh hm, maybe my MAC did change when I IPv6'd and I didn't know it
14:54<straterra>It did
14:54<kenyon>I guess that's what the reboot was for
14:59<fo0bar>woot, now both my linode and my dedicated colo are native ipv6
14:59<fo0bar>(colo for about a year, linode had been running an HE tunnel until now)
15:00-!-Cromulent [] has quit [Remote host closed the connection]
15:01<fo0bar>that was eerie
15:01<fo0bar>my linode is named :)
15:02-!-maushu [] has joined #linode
15:02-!-digilink [] has joined #linode
15:03<tjfontaine>fo0bar: btw, congrats on new employment
15:03-!-Boohemian [~Boohemian@] has joined #linode
15:04-!-Dreamer3 [] has quit [Quit: Computer has gone to sleep.]
15:04<Daevien>you are finally paying him now tj? :p
15:05<tjfontaine>I don't have anywhere near the cash for what he deserves :)
15:05<fo0bar>tjfontaine: thanks. I'll probably be moving to the Philly area in about a year, within stalking distance of Linode HQ :)
15:05*Zr40 wonders if the visibility changed now that it started raining quite heavily
15:05<Zr40>!wx EHEH
15:05<+linbot>Zr40: [metar] OBS at EHEH: 68.0F/20C, visibility 6000 miles, wind 17.26 mph (altimeter: 29.972932745) [EHEH 281856Z AUTO 31015G26KT 6000 +TSRA FEW014 FEW015CB SCT055 BKN060 20/20 Q1015 WHT BLU TEMPO YLO 14008KT 9999 BKN120 TEMPO 24018G38KT 2000 TSRAGR SCT035 BKN045CB]
15:05<tjfontaine>fo0bar: if not the HQ a few of its employees
15:05<Zr40>1600 -> 6000... not a decrease
15:05<straterra>!wx IND
15:06<+linbot>straterra: [metar] IND: not found
15:06<Zr40>!wx KIND
15:06<+linbot>Zr40: [metar] OBS at KIND: 78.8F/26C, visibility 10 miles, wind 10.36 mph (altimeter: 30.02) [KIND 281854Z 31009G17KT 10SM FEW045 SCT180 26/14 A3002 RMK AO2 SLP158 T02610144]
15:06<@caker>!wx kacy
15:06<+linbot>caker: [metar] OBS at KACY: 75.2F/24C, visibility 1 miles, wind 11.51 mph (altimeter: 29.9) [KACY 281835Z 03010KT 1 1/2SM -RA BR FEW002 OVC046 24/23 A2990 RMK AO2 P0021 ]
15:06<Zr40>1 mile... not a lot
15:06-!-pigdude [] has quit [Quit: leaving]
15:07<squircle>!wx cwwb
15:07<+linbot>squircle: [metar] OBS at CWWB: 78.8F/26C, visibility CWWB 281800Z AUTO 26009G17KT 26/15 RMK AO1 PK WND 28018/1743 SLP062 T02590148 55009 miles, wind 10.36 mph (altimeter: ) [CWWB 281800Z AUTO 26009G17KT 26/15 RMK AO1 PK WND 28018/1743 SLP062 T02590148 55009]
15:07-!-Rezt [] has joined #linode
15:08<squircle>i half-expected to see (altimeter: null)
15:08<Zr40>what does altimeter indicate?
15:08<squircle>air pressure
15:08<squircle>for airplanes to calibrate their altimeters
15:09<squircle>I just said I expected to see null because I checked the weather at a nautical weather station, not one meant for planes
15:10<squircle>(29.92 inHg = 1 atm)
15:15-!-orpheus [] has joined #linode
15:16-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
15:16-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
15:17-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
15:18-!-Rezt [] has left #linode []
15:18-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
15:18<Daevien>!wx CYQI
15:18<+linbot>Daevien: [metar] OBS at CYQI: 71.6F/22C, visibility 15 miles, wind 5.75 mph (altimeter: 29.97) [CYQI 281900Z 18005KT 15SM FEW020 OVC250 22/16 A2997 RMK SC1CI2 TR SC CC ASOCTD WND DCTN ESTD SLP150]
15:19-!-mario [~mario@] has joined #linode
15:21-!-chmac [] has quit [Quit: Leaving]
15:21-!-mario [~mario@] has quit []
15:22-!-JSharp [] has joined #linode
15:23<hawk>!wx ESSB
15:23<+linbot>hawk: [metar] OBS at ESSB: 75.2F/24C, visibility 9999 miles, wind 5.75 mph (altimeter: 30.150112643) [ESSB 281850Z 21005KT CAVOK 24/16 Q1021]
15:23-!-xt3mp0r [~xt3mp0r@] has joined #linode
15:23-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
15:25-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
15:25-!-laser` [] has joined #linode
15:25-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
15:27<Zr40>!wx EHEH
15:27<+linbot>Zr40: [metar] OBS at EHEH: 69.8F/21C, visibility 9999 miles, wind 9.21 mph (altimeter: 29.913872779) [EHEH 281925Z AUTO 09008KT 9999 -TSRA FEW015CB SCT063 BKN090 21/21 Q1013 RERA BLU BLU TEMPO YLO 14008KT 9999 BKN120 TEMPO 24018G38KT 2000 TSRAGR SCT035 BKN045CB]
15:27<DrJ>!wx abcd
15:27<+linbot>DrJ: [metar] ABCD: not found
15:27<DrJ>whoever runs that should like change it to zip code or something
15:28<laser`>Tricky, METAR is weather stations from airport
15:28<Zr40>hawk, DrJ: actually, it's the ICAO airport code
15:28<squircle>so much easier that way
15:28<laser`>!wx EGCN
15:28<squircle>hawk: that list is really useful, thanks!
15:28<+linbot>laser`: [metar] OBS at EGCN: 62.6F/17C, visibility 9999 miles, wind 8.06 mph (altimeter: 30.031992711) [EGCN 281850Z 33007KT 300V360 9999 FEW019 BKN041 17/11 Q1017]
15:28<hawk>squircle: np
15:29-!-voker57__ [] has quit [Ping timeout: 480 seconds]
15:29<Zr40>hawk: heh, not often that it's warmer in sweden than in the netherlands :)
15:29-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
15:30-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
15:31-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
15:31-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
15:33-!-stafamus [~stafamus@] has joined #linode
15:33<Daevien>hmm, filled the 750gb download drive on my home server, i think it's time i actually weed through the junk in there, 0 bytes free :p
15:34<hawk>Zr40: I suppose not, it's been a hot day here anyway
15:34-!-voker57__ [~voker57@] has joined #linode
15:35<Zr40>hawk: it was ~28C earlier today here
15:35<Zr40>then this pretty nasty storm paid a visit
15:36-!-skegeek [] has joined #linode
15:37<hawk>Zr40: I think it was over thirty here
15:37<skegeek>Do I need to install AIDE for Xen or would that be for use by Linode's system?
15:39<tjfontaine>if you're using pv kernel you shouldn't need the -xen package I don't think it would be necessary
15:40<tjfontaine>but I don't know the software that well
15:40<tjfontaine> suggests that it is suitable for the DomU (which is your instances classification)
15:40<skegeek>I'm not sure which kernel I'm using...
15:41<tjfontaine>probably Latest 2.6 in the manager
15:41<skegeek>uname -a returns: Linux [hostname] #1 SMP Thu Apr 21 20:08:53 UTC 2011 i686 i686 i386 GNU/Linux
15:42<skegeek>It would be really nice if a 512 could handle supporting remote X11 as well as a few sites, but oh well.
15:43<Daevien>eh, it should be able to. but why do you want X on a remote vps?
15:43<skegeek>It would provide gui administration
15:44-!-synapt [] has quit [Read error: Connection reset by peer]
15:44-!-synapt [] has joined #linode
15:45<Daevien>it will b ebetter if you learn to use cli admin and have a better handle on how stuff works without adding an additional security holes
15:46<tjfontaine>skegeek: in other words, most of us find it inconvenient to do anything but work from a command line
15:47<skegeek>I have been learning how things work, but there are times when I need to be able to do something promptly and learning isn't always so quick.
15:47<tjfontaine>skegeek: it's not been my experience that you find useful GUI administration tools
15:47<tjfontaine>unless the G you refer to is really a webui
15:48<tjfontaine>at which point I still question its actual usefulness, but appreciate its utility
15:48<skegeek>I mean for something like adding a user for file-transfer.
15:48<Daevien> & you've already foudn this channel. linode is unmanaged (so that means staff doesn't have to help you outside of getting the basics going) but staff & other users will help if they are around and can
15:49<tjfontaine>if you want a bit more of a "guided" experience administering your node you could use a control panel
15:52<skegeek>And, I have also sometimes thought it would be both beneficial (and rather cool) to have a centralized system for use between people I work with...which aren't tech savvy. Might have somewhat to do with the inner urge to feel a bit "uber cool" for what I have done/setup.
15:54-!-Cromulent [] has joined #linode
15:54<skegeek>Out of curiousity, would a couple of people using a web os be more or less secure than remote X11?
15:55<Zr40>what's a 'web os'?
15:55<pronto>it's like a bagel
15:55-!-userme [] has quit [Quit: userme]
15:55<skegeek>I mean a web-based desktop, like ChromeOS for example.
15:55<pronto>its good and all, but with a massive hole in it
15:56-!-disinpho [] has joined #linode
15:56<skegeek>Well, ChromeOS minus having to install on the local machine.
15:57<Zr40>how would you access such a thing?
15:57<skegeek>It would run inside a web browser like any other web application.
15:58<Zr40>so... you would access a remote web browser using a web browser.
15:58<Zr40>I don't see the gain of that
15:58<skegeek>The GUI's would be like AJAX front-ends or something.
15:59<skegeek>No no. I'd use a web browser to access the web-based OS just as you do a web app, since technically it'd be a web app anyway.
15:59<Zr40>what would you be able to do in this web-based OS?
16:00<skegeek>I'm not talking about a new idea, I'm talking about a private ChromeOS type of thing, hosted on vps.
16:00<Zr40>ChromeOS is a bare OS with a browser in it
16:00<hawk>But... ChromeOS is a means of accessing web applications, isn't it?
16:00<skegeek>Yeah, because everything else is web based...
16:01<Zr40>exactly. so any regular browser would be able to do that
16:01<Zr40>using a regular browser to connect to a web-based ChromeOS-like thing just does not make sense
16:01<skegeek>Yeah, I didn't mean special browser, I just meant a private vps hosted system with any apps needed.
16:01-!-Boohemian [~Boohemian@] has quit [Quit: Computer has gone to sleep.]
16:03<Zr40>skegeek: what kind of apps?
16:03<skegeek>Anyway, so I do -not- need to install the Xen version of AIDE?
16:03<Zr40>need? no, you don't need to
16:05<skegeek>Zr40: Anything...from project management and customer relationship manager to email client and web-based file-transfer application. Whatever type of application needed.
16:05<swaj>I <3 BizSpark :P
16:05<skegeek>Why is there a Xen specific version anyway?
16:05<swaj>free production licenses ftw
16:05<skegeek>Isn't BizSpark free software from Microsoft?
16:06<navi>As a student, I get to use DreamSpark. I like that. It gives me free Windows Server licenses. If I ever need windows, I can spin one up in a VM.
16:08<Cromulent>shame microsoft dropped the version of VS you got with dreamspark - you used to get the middle version of VS with the 2008 version now you get the low end version of 2010
16:08<Cromulent>even though they are both called the Pro version - can't fault microsoft's marketing department
16:10-!-lsabota [] has quit [Quit: leaving]
16:11-!-Dreamer3 [~dreamer3@] has joined #linode
16:11-!-cereal is now known as cereal|Away
16:12-!-pygi [] has joined #linode
16:12<navi>Yeah, true
16:12<pronto>i agree
16:13<skegeek>Which log view software do you guys recommend? To me Webalizer looks like a 1990's software; Piwik looks pretty good, but I couldn't figure out how to get the code to embed into a site.
16:13<pronto>if you have the monies , splunk is pretty good
16:13*nDuff uses (and adores) logstash
16:14<bob2>oh, web analytics
16:14<nDuff>bah re: splunk -- no support for hierarchial parsing
16:14<skegeek>By the way, should I be installing a separate instance of such software for each virtual host I setup?
16:14<nDuff>I want to be able to parse my logs into _fields_, and do queries against those fields, damnit -- full-text searching just isn't enough.
16:14<bob2>nDuff: ^ is that what you mean by hierarchal parsing?
16:14<nDuff>also, bah re: loggly following splunk's lead (yes, I know it's the same people)
16:15<nDuff>bob2, yes -- with logstash, I can break things into fields, and break those down into subfields, and query against each (sub)field separately.
16:17<navi>Ugh, I hate saying this but...
16:17<nDuff> if I say I want to search for responses with HTTP status 500, I don't get log entries showing number-of-bytes-returned 500, or user ID 500, or [...]
16:17<navi>Screw this, I'm moving back to Excel
16:17-!-jake [] has quit [Ping timeout: 480 seconds]
16:18<navi>That is still one area I like MS more for.
16:18<nDuff>navi, ...what area?
16:18<priyesh>any good resources with a basic IPtables config?
16:18<priyesh>to build upon
16:18<navi>nDuff: Spreadsheeting
16:18<nDuff>priyesh, I strongly, strongly suggest shorewall
16:19<navi>nDuff: Well, any more complicated spreadsheets. I find, for instance, apple numbers to be lacking with some of the charting options
16:19-!-jake [] has joined #linode
16:22<bob2>shorewall <3
16:24-!-joshdotsmith [] has joined #linode
16:27-!-TIBS02 [lemfpomwe@] has joined #linode
16:27-!-jxpx777 [~jxpx777@] has joined #linode
16:31-!-TIBS01 [lemfpomwe@] has quit [Ping timeout: 480 seconds]
16:33-!-lsabota [] has joined #linode
16:36-!-piney0 [] has joined #linode
16:37-!-Jesta [] has joined #linode
16:38<Zr40>OHGOD BEES!
16:39<swaj>priyesh: If you're using debian or ubuntu, I recommend ufw. It's iptables without the headache :)
16:40<priyesh>swaj: i'm using debian
16:40<swaj>sudo apt-get install ufw && man ufw
16:40<nDuff>ehh. shorewall has the advantage that it'll grow with you to pretty much however complex a configuration you might need
16:40<nDuff>Not so sure about ufw in that regard.
16:40<swaj>ufw = OpenBSD-style ipfw rules that translate into iptables rules for you.
16:40<nDuff>...really, that powerful?
16:40<swaj>ufw is just syntactic sugar on top of iptables, so it'll go wherever you want it to.
16:40*nDuff might need to give it another look.
16:41<AviMarcus>ufw is pretty cool stuff.
16:41<priyesh>on a new debian install, what rules would i want to apply?
16:41<priyesh>from scratch
16:41<AviMarcus>Yeah, ufw default deny, then ufw allow ports...
16:41<swaj>priyesh: by default, ufw denies everything. So you just want to allow specific stuff
16:41<priyesh>swaj: what kind of stuff? ssh, http, ?
16:41<AviMarcus>swaj, I don't think default is actually deny
16:41<swaj>priyesh: so, I'd just start with one rule: ufw limit ssh
16:42<AviMarcus>limit ssh?
16:42<swaj>it was when I installed it
16:42<swaj>yeah it throttles, so if you get someone thrashing your SSH port, iptables will reject them :P
16:42<nDuff>swaj, ...well, sure, _everything_ is syntax sugar on top of iptables; the question is whether they allow the full expressive power.
16:42<AviMarcus>hmm that's cool
16:42<swaj>nDuff: in my experience, it allows pretty much everything I've wanted to do.
16:42<AviMarcus>nDuff, they don't have to.
16:42<AviMarcus>You can still do iptable rules..
16:43<AviMarcus>just make sure it's higher than the deny
16:43<nDuff>swaj, ...looking at ufw's changelog, there are a lot of features I use regularly in shorewall which ufw didn't support prior to 10.04 -- so it looks to be still an evolving thing.
16:43<swaj>it is, sure
16:43<hawk>chesty: Let's press it again!
16:43<swaj>but it's damn nice :)
16:44<Obsidian|server>swaj: it's default to deny inbound, but not outbound
16:44*Obsidian|server has it denying outbound by default as well as inbound
16:44<swaj>priyesh: so my basic rules are like: ufw limit ssh/tcp, ufw allow http/tcp, ufw allow https/tcp
16:44<swaj>Obsidian|server: yeah I don't like default deny on outbound, makes no sense to me.
16:45<Obsidian|server>I do it because I'm paranoid about my system being hijacked
16:45<swaj>priyesh: all other traffic (inbound) will get denied.
16:45*Obsidian|server shrugs
16:45<Obsidian|server>priyesh: if you use ftp/ftps, you'll need to open up that too
16:45-!-RoosterJuice [] has joined #linode
16:45<swaj>yeah, although who does that when you have SSH? :P
16:45<Obsidian|server>personally, you're best off using ftps or sftp
16:46<swaj>SCP/SFTP ftw
16:46<Obsidian|server>and ftps is a pita
16:46<Obsidian|server>^ what swaj said
16:46-!-TIBS02 [lemfpomwe@] has quit [Ping timeout: 480 seconds]
16:47<priyesh>so would the following (pure iptables) be a good starter:
16:47<priyesh>Chain INPUT (policy ACCEPT)
16:47<priyesh>target prot opt source destination
16:47<priyesh>ACCEPT tcp -- anywhere anywhere tcp dpt:www
16:47<priyesh>ACCEPT tcp -- anywhere anywhere tcp dpt:https
16:47<priyesh>DROP tcp -- anywhere anywhere
16:47<+linbot> <-- paste here, not in the channel
16:47<priyesh>sorry, !p
16:47<swaj>for your input chain, that looks fine.
16:47-!-flowbee [] has quit [Read error: Connection reset by peer]
16:48<Obsidian|server>jquery is thinking that an element's height, when not visible, is greater than when it is visible
16:48<retro|blah>priyesh: How are you going to ssh in, though?
16:48<swaj>LISH :P
16:48<Obsidian|server>it's adding 30px -_-
16:48<swaj>although I can't use LISH fulltime because I need > 80 character width :P
16:48<swaj>(or whatever the default term width is)
16:49<swaj>I need maximum girth for my term window
16:49<retro|blah>210 columns \o/
16:49<priyesh>so this is fine:
16:50<swaj>I'd change the ssh rule from ACCEPT to LIMIT
16:50<priyesh>do i need to touch OUTPUT or INPUT/udp?
16:50<swaj>not for ssh/http/https -- they don't operate on udp
16:50<priyesh>what does limit do?
16:51<swaj>it tells iptables to reject people that try and flood your SSH port with login attempts
16:51<priyesh>so should I do INPUT: DROP udp
16:51<priyesh>if i'm not using udp
16:51<hawk>Not with login attempts specifically, but with new connections, I guess?
16:51<swaj>your DROP rule should be "ip" for protocol, not just "tcp"
16:51-!-tharkun [~0@] has joined #linode
16:51<swaj>hawk: yeah
16:51<swaj>the "ip" prot covers both tcp and udp
16:51<hawk>Or maybe just set the policy to DROP
16:52<hawk>instead of having a special rule for it
16:52<swaj>that would be even better :P
16:52-!-minecraftfan [~minecraft@] has quit [Remote host closed the connection]
16:52<priyesh>but that would stop pings too
16:52<swaj>forget the DROP rule completely and change the INPUT chain's default policy to DROP.
16:52<hawk>priyesh: So add a rule that allows those
16:52<swaj>then you can put in an ACCEPT for icmp pings
16:53-!-minecraftfan [~minecraft@] has joined #linode
16:53<tharkun>caker: i was looking at the library a few moments ago an i saw that you have an exim config tutorial to work as a nullmailer. Why not do a tutorial about nullmailer (the program) or some other version. It would save a lot of ram to many people and use their linode more efficiently
16:54<tharkun>an example would be msmtp
16:54<priyesh>swaj: how do i do that :P
16:55-!-Protolich [] has joined #linode
16:56<priyesh>swaj: i meant change the default policy
16:57<swaj>sudo iptables -T FILTER -P INPUT DROP
16:57<swaj>should do the trick
16:59<Cromulent>linode really should offer a SAN(s) that people can rent space on
17:00<tjfontaine>should they really
17:00<Cromulent>well I would like to see it as an added service - lets put it that way
17:00-!-array [danny@2600:3c01::f03c:91ff:fe96:db42] has left #linode []
17:00-!-array [danny@2600:3c01::f03c:91ff:fe96:db42] has joined #linode
17:00-!-mode/#linode [+o array] by ChanServ
17:00-!-dajhorn [] has quit []
17:01<priyesh>is it fine to accept all OUTPUT traffic?
17:01-!-redgore [~redgore@] has quit [Quit: Leaving]
17:01-!-kenichi [] has joined #linode
17:01<Cromulent>I don't - best just to allow what you know you need
17:01<Cromulent>which reminds me I need to accept the NTP ports again :p
17:03<hawk>It's just 123, isn't it?
17:03<priyesh>Cromulent: which ports do i need to allow in OUTPUTs
17:03-!-flowbee [] has joined #linode
17:04<Cromulent>priyesh: I have 22/tcp 53/udp 80/tcp and 443/tcp enabled on mine and soon to add NTP as well
17:04<Cromulent>you'll need more if you a vpn or email etc
17:04<Cromulent>you run*
17:04-!-bonjurkes [] has quit [Quit: ajax IRC Client]
17:04<Cromulent>also have 22/tcp rate limited
17:04<priyesh>Cromulent: DNS uses 53/udp?
17:05<hawk>and 53/tcp to some extent
17:05<hawk>But primarily udp
17:06<priyesh>what about if i'm running an irc client?
17:06<Cromulent>6667 perhaps?
17:06<bob2>then you need to allow outbound traffic
17:06<Cromulent>no idea about irc really
17:06<bob2>and not be a moron and DENY everything
17:06<bob2>or ident will hang
17:07<hawk>bob2: Indeed... better reject the ident port
17:08-!-JSharp [] has quit [Quit: Leaving]
17:10<hawk>Might want to just have reject as the policy even... I guess that comes down to your preference and philosophy....
17:10-!-linville [] has quit [Quit: Leaving]
17:11<Kyhwana>Also make sure to allow ICMP(v6 too)
17:11-!-copperx [] has joined #linode
17:11<bob2>pmtu is for suckas
17:11-!-Dreamer3 [~dreamer3@] has quit [Ping timeout: 480 seconds]
17:15-!-Dreamer3 [] has joined #linode
17:16-!-minecraftfan [~minecraft@] has quit [Remote host closed the connection]
17:17-!-lakin [] has quit [Quit: Ex-Chat]
17:17-!-minecraftfan [~minecraft@] has joined #linode
17:18<hawk>Well, if you allow RELATED that solves a lot v4 without outright allowing all icmp. For v6 there are some other things you'd want in addition, though.
17:19-!-Protolich [] has left #linode []
17:20<priyesh> - what's wrong with this
17:20<priyesh>can't access webserver
17:22-!-squircle [] has quit [Quit: this is not the default quit message]
17:22<akerl>priyesh: And you can access without those rules?
17:23<hawk>priyesh: I think you want to allow ESTABLISHED,RELATED
17:24<swaj>imo just allow everything out
17:24<swaj>no point in blocking outbound
17:24-!-marcopkb [] has joined #linode
17:24<priyesh>swaj: i'd agree tbh
17:25<priyesh>swaj: what about FORWARD?
17:26<swaj>I believe that has to do with your server acting as a router or proxy, and forwarding traffic to another device
17:26<swaj>so you can either leave it alone, or probably do a default drop if you're paranoid.
17:26<priyesh>swaj: so i could just DROP it and forget about it
17:26<priyesh>swaj: okay :)
17:29-!-xt3mp0r [~xt3mp0r@] has joined #linode
17:29-!-Zr40 [] has quit [Quit: leaving]
17:29-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
17:30-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
17:31-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
17:32-!-jared0x90 [] has quit []
17:36-!-Jippi [] has quit [Quit: Jippi]
17:37-!-tempesta [] has quit [Ping timeout: 480 seconds]
17:37-!-Cromulent [] has quit [Remote host closed the connection]
17:39-!-JediMaster [] has joined #linode
17:40-!-tempesta [] has joined #linode
17:41-!-DrJ [] has quit [Quit: Bye]
17:41<+linbot>New news from wiki: User:Loadavg <> || User:Loadavg <>
17:41<priyesh>ssh doesn't require 22/udp, does it?
17:41<retro|blah>Don't think so
17:42<priyesh>well this is what i've ended up with
17:42<priyesh>i've learnt a *lot* from working with iptables
17:44*JediMaster ponders cluster design
17:45-!-RoosterJuice [] has quit [Quit: This computer has gone to sleep]
17:45<JediMaster>bleh, if I ditch the secondary mysql server in this setup and make it a webserver I'll get 50% more performance but loose the redundancy of a second db server
17:46-!-DrJ [] has joined #linode
17:46<swaj>doh wrong window
17:47<JediMaster>anyone know if the backup service only backs up ext3 partitions?
17:47<JediMaster>as I've got 150GB raw DRBD/GFS2 partitions I'd like backing up too
17:47<Knorrie>just try?
17:48<Knorrie>i think it's block-level snapshot dyncing
17:48<chesty>it backs up on a file level
17:48<Knorrie>to the backup-location
17:48<priyesh>with iptables, what's the difference between DROP and REJECT?
17:48<akerl>priyesh: Drop is silent, reject isn't
17:48<ajmitch>JediMaster: the FAQ says it needs to be able to mount the disk images on the host
17:48<priyesh>akerl: understood
17:48<swaj>DROP just drops the connection silenty without informing the end client. REJECT will tell the end client that the connection is refused.
17:48<ajmitch>currently just ext2/3
17:49-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
17:49-!-bbeausej1 [] has quit [Quit: Leaving.]
17:49-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
17:49<priyesh>swaj: so any specific circumstances I should use REJECT?
17:50<deejoe>if you want to hurt the remote hosts feelings
17:50<bob2>unless you have a reason, don't DENY stuff
17:50<swaj>I honestly don't use reject because I'd rather just DROP and let them time out
17:50-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
17:50<Knorrie>priyesh: e.g. if you want to block outgoing traffic, you generally want the process trying to know directly it cannot do that, instead of hanging some tiem
17:51<hawk>drop is kind of a way of slightly punishing the one connecting...
17:51-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
17:51<swaj>but for incoming traffic that you haven't explicity allowed, DROP makes them wait until they time out instead of immediately informing them that the connection was rejected.
17:51-!-cereal|Away is now known as cereal
17:51<Knorrie>tarpit is even more fun
17:52<Knorrie>escpecially for incoming tcp/25 on non-mailservers
17:52<swaj>so yeah, I'd say REJECT on outbound (if you're filtering outbound, which I still think is silly if you trust your users), and DROP in inbound.
17:53<bob2>if you're running php stuff, I'd suggest blocking some outbound ports
17:53<Knorrie>some trusted still install joomla
17:53<priyesh>swaj: makes sense :) ( -> not filtering outbound)
17:53<bob2>so when it gets pwned it can't connect to a botnot or an irc bot control channel
17:53<Knorrie>bob2: yup, that's what i meant
17:53<bob2>er botnet
17:53<Kyhwana>you can also block outgoing/incoming based on PID or user
17:53<bob2>botnot sounds like a robot crime
17:54<Knorrie>Kyhwana: only outgoing
17:54<swaj>and now you know why I use ufw :P
17:54<swaj>iptables gives me a headache :P
17:54<Knorrie>iptables is fun
17:56-!-xt3mp0r [~xt3mp0r@] has joined #linode
17:56<+linbot>New news from forums: I feel blown in Customer Testimonials <>
17:57-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
17:57-!-DrJ [] has quit [Read error: Connection reset by peer]
17:57-!-DrJ [] has joined #linode
17:57-!-Farkle [] has joined #linode
17:57-!-DrJ [] has quit [Read error: Connection reset by peer]
17:57-!-disinpho [] has quit [Ping timeout: 480 seconds]
17:57<hawk>pf is fun... iptables takes a bit of a masochistic side for it to be fun, I think
17:58<swaj>wow he feels blown
17:58<swaj>he loves Linode so much, that it blew him.
17:58-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
17:58<swaj>caker must have put out
17:59<Kyhwana>Knorrie: er, yeah, you can block based on user account for outgoing too. (I know, I do this)
17:59-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
17:59<bob2>-> #teens
17:59-!-Farkle is now known as DrJ
18:00-!-Jesta [] has quit [Remote host closed the connection]
18:00-!-BaldwinKoo [] has quit [Read error: Connection reset by peer]
18:00-!-disinpho [] has joined #linode
18:00-!-BaldwinKoo [] has joined #linode
18:01<DrJ>I'd like to file a complaint against linode
18:01<hawk>Oh noes
18:02<DrJ>I've been using linode now for like 8 months and not once have they screwed me ... I mean, I can get screwed by every other provider out there (almost on a daily basis) ... and they don't even charge for it
18:02<swaj>apparently they blew one guy: <>
18:03-!-lunks_ [~lunks@] has joined #linode
18:03<Knorrie>Kyhwana: that's what i said :) *only* outgoing
18:03-!-amar__ [] has quit [Ping timeout: 480 seconds]
18:04<bob2>omg the tor thread continues
18:04<hawk>What now?
18:04<Knorrie>grab a beer?
18:05<bob2>sign up for google one!
18:06<tonyyarusso>You mean Google+$!@%&?
18:07<pharaun>what about google two?
18:07<bob2>oh right
18:08<Knorrie>google zero, no artificials added
18:09-!-lunks [] has quit [Ping timeout: 480 seconds]
18:09-!-lunks_ is now known as lunks
18:10-!-priyesh [] has quit [Ping timeout: 480 seconds]
18:10-!-priyesh [] has joined #linode
18:10<bob2>free linode beer
18:10<akerl>"i feel blown"? Really? The whole english language, and that's what best compares to linode's service?
18:11<bob2>I assume it's a non-native speaker and they forgot "away"
18:11<akerl>I hope :p
18:11<tjfontaine>I'm ok if they feel like they just got a great blowjob
18:11<bob2>it is an odd idiom
18:12<hawk>Yeah, sounds like high praise... if a bit crude...
18:13<tjfontaine>maybe they're a pornstar
18:13-!-bigjocker [~ngranek@] has quit [Remote host closed the connection]
18:14<JediMaster>hmm, doesn't drbd work with more than 2 nodes? =/
18:14-!-djg320 [] has joined #linode
18:14<tjfontaine>JediMaster: yes but you have to layer, it works in pairs
18:14<bob2>how are you going to brbd that
18:16<JediMaster>tjfontaine: like that?
18:16<JediMaster>say in that example, I have the two machines alice and bob, already dual primaries...
18:17<JediMaster>should the ip in that "stacked-on-top-of r0" section be one of the ips of those first two machines?
18:17<JediMaster>all three are on the same subnet here
18:17<tjfontaine>is this your first experience with drbd?
18:18<JediMaster>the last few days, yup
18:18<JediMaster>I've got dual-primaries with corosync and GFS2 working
18:18<tjfontaine>but you are confident in your exisiting non-layered setup?
18:19<JediMaster>I've copied several GB of data to it with the second one down and brought it back up and had it all sync up to date in about 25 sec
18:19<JediMaster>and working both ways
18:19<tjfontaine>well, the ip rules for layering work the same as the base set
18:20<tjfontaine>so the stacked-on-top-of is generally considered the primary
18:20<JediMaster>so do I just pair up one of the two primaries with the new one?
18:20<tjfontaine>but I'm not sure I would want to layer and dual primary
18:20<JediMaster>and can you dual-primary in the layered setup too?
18:21<JediMaster>was hoping I could expand the FS clustering beyond 2 machines
18:21<tjfontaine>I've not done stacking in dual primary, and there are other file systems that are designed for these types of operations
18:21<JediMaster>I wrongly assumed you could just add another device to the resource and sync it
18:22<JediMaster>tjfontaine: drbd isn't a file system really
18:22<tjfontaine>I know what drbd
18:22<JediMaster>well then =)
18:22<tjfontaine>I'm saying, that there are filesystems designed for clustering that do not rely on block layer mapping
18:22<JediMaster>such as?
18:22<tjfontaine> which is #ceph here
18:22-!-jxpx777 [~jxpx777@] has quit [Quit: Mac is sleeping…]
18:23<JediMaster>is that part of latest linux?
18:23<tjfontaine>excersize for the reader :P
18:23*JediMaster does a make menuconfig
18:23<tjfontaine>Distributed file system. An dynamic cluster of metadata servers create and manage and file system hierarchy, providing POSIX file system access via the Ceph file system client in recent Linux kernels or via a FUSE driver.
18:24<hawk>So at least the client bits
18:24<tjfontaine>anyway, it's all an excersie for the reader, time for me to go home
18:25<JediMaster>can't see it in menu config on
18:25<tjfontaine>rbd is there?
18:26<hawk># CONFIG_CEPH_LIB is not set # CONFIG_CEPH_FS is not set
18:26<tjfontaine>no just rbd, anyway Kernel version 2.6.34 already includes the Ceph kernel client.
18:26<hawk>from my running kernel's config.gz
18:26<JediMaster>well that's interesting, I had to roll my own on linode to get OCFS2 and GFS2 support
18:27<tjfontaine>of course
18:27<tjfontaine>you could have alternatively used the distros mechanism
18:27<JediMaster>I see it's not set in my .config
18:27<JediMaster>in that it's there, just not set
18:27<JediMaster>but can't find it under file systems
18:27<tjfontaine>probably missing a dep
18:27*tjfontaine leaves work
18:28<JediMaster>this would bypass the need for drbd entirely and I guess the custom kernel too
18:29<JediMaster>ubuntu has ceph-fuse as well as the client tools
18:29<hawk>Aha.. with fuse maybe...
18:29-!-Dreamer3 [] has quit [Quit: Computer has gone to sleep.]
18:31<djg320>I'm seeing more than 100mb reserved for kernel on a Lin 2048
18:32<swaj>which kernel? and are you on 32-bit?
18:32<djg320>Mem: 1980280k total, 63596k used, 1916684k free, 1196k buffers
18:32<djg320>64 bit latest paravirt
18:32<swaj>yeah that's why
18:32<swaj>32-bit latest paravirt would be much better
18:32<swaj>64-bit kernel reserves a lot of RAM
18:32<swaj>it got a lot better in
18:33-!-jamescollins [~jamescoll@] has joined #linode
18:33<hawk>Well, both 32 and 64 bit started doing it in .39... 32 bit has been fixed, 64 bit not so much
18:33<djg320>Hmm, don't like the idea of nuking and paving the linode
18:33<djg320>Wait it out?
18:34<hawk>I suppose you could run the .38 kernel and hope x86-64 gets some sort of fix for it too
18:34-!-AviMarcus [~avi@] has quit [Quit: Ex-Chat]
18:35<hawk>I have no idea what the security implications may be
18:36<djg320>Nothing like a Linode boot time.
18:37-!-jmartinez [] has joined #linode
18:37<djg320>Gave me 66mb of ram back as usable
18:37<djg320>Mem: 2047132k total, 57964k used, 1989168k free, 1184k buffers
18:37-!-tempesta [] has quit [Ping timeout: 480 seconds]
18:38-!-tempesta [] has joined #linode
18:38<djg320>Linux web2 2.6.38-x86_64-linode17 #1 SMP Thu Apr 28 22:18:47 UTC 2011 x86_64 GNU/Linux
18:39-!-ktabic [] has quit [Quit: I'm a professionally trainined computer scientist. That is to say, I am poorly educated]
18:41<priyesh>to run a script at bootup, do i add it to init.d then update-rc.d _or_ put it in rc.local
18:41-!-skegeek [] has quit [Read error: Connection reset by peer]
18:42-!-skegeek [] has joined #linode
18:43<jmartinez>I am trying to configure nginx + php-fpm. When i try to start the php-fpm service it says the port(9000) is already in use.
18:43<bob2>sudo netstat -plnt | grep 9000
18:43<jmartinez>tcp 0 0* LISTEN 2452/php5-cgi
18:44<swaj>seems you've got php5-cgi installed
18:44<bob2>and an init script for it
18:44<swaj>sudo apt-get remove php5-cgi :P
18:44<swaj>php5-fpm replaces php5-cgi, you don't want both :)
18:44<jmartinez>when i look at a phpinfo page it lists this under server API: CGI/FastCGI
18:45<jmartinez>Dont I want to see php-fpm?
18:45<swaj>yeah you're using traditional fastcgi, which is fine.. but php-fpm is better
18:45<bob2>remove php5-cgi and see if that stops the boave
18:45-!-Dreamer3 [] has joined #linode
18:45<swaj>remove php5-cgi, then start your php5-fpm on that port, and your info.php should show FPM/FastCGI
18:47-!-Cromulent [] has joined #linode
18:49<JediMaster>nginx + php-fpm rocks
18:50<swaj>aye, as much as PHP can rock that is :P
18:50<JediMaster>it rocks enough
18:50<swaj>wee IIS 7.5 FTP configured... had to require SSL to make myself feel better :P
18:50-!-lunks [~lunks@] has quit [Quit: lunks]
18:50<swaj>I may firewall it so only I can get in...
18:51<swaj>actually yeah, I'm gonna do that
18:51<jmartinez>Now I get a 502 bad gateway error when trying to view thr phpinfo page
18:52<bob2>== read your nginx logs
18:52<swaj>so make sure php-fpm is running... make sure nginx is set to forward to the right port
18:53<swaj>that's nginx telling you that it couldn't find your FastCGI server (php-fpm)
18:53<akerl>jmartinez: At the moment, what is the current state of 1) nginx, 2) phpfpm, 3) fastcgi?
18:54<swaj>there is no "fastcgi" program in the nginx + php-fpm combo :P
18:55<akerl>tcp        0      0*               LISTEN      2452/php5-cgi
18:55<swaj>that's php5-cgi, that's not php-fpm
18:55<swaj>two very different things
18:55<akerl>Sure. s/fastcgi/php5-cgi/
18:55<swaj>php5-cgi shouldn't be installed if you're using php-fpm
18:55<swaj>they're mutually exclusive
18:55<akerl>The relevant factor is the state of "thing that is not phpfpm, and interferes with"
18:56<swaj>though he's getting a 502 which would imply php5-cgi is likely gone.
18:56<swaj>or at least not running
18:56<swaj>I think I made my php-fpm listen on a unix socket
18:57<swaj>not sure if that's default or not
18:57<akerl>Yes, which is why I expect that nginx is running, php5-cgi is gone, and phpfpm is maybe on, but nginx isn't configed to use it
18:58<jmartinez>I am not getting anywhere. (Newb) What is the first step here guys?
18:58<akerl>jmartinez: Answer my question above?
18:58<swaj>jmartinez: first step is to make sure php-fpm is running.
18:58<jmartinez>how do I check?
18:58<akerl>ps aux?
18:58-!-nicinabox [] has quit [Quit: nicinabox]
18:58<bob2>ps aux | grep -i php
18:59<bob2>sudo netstat -plnt | grep 9000
18:59<akerl>First step is to ensure that php-fpm is running, and no other php junk
18:59<swaj>run the two commands bob2 typed, and paste the output in a pastie
18:59<+linbot> <-- paste here, not in the channel
18:59<akerl>Then, see if phpfpm is using a socket, or a port.
19:00<akerl>Then, make sure that nginx is configured to use that connection method
19:00<swaj>I think by default, php-fpm listens on
19:02<jmartinez>here is the htop results:
19:02-!-tunabananas1 [] has joined #linode
19:02<akerl>jmartinez: `netstat -an | grep 9000`
19:02<akerl>Is phpfpm there?
19:02<swaj>so php-fpm is running
19:03<swaj>okay and it's listening on the right port
19:03<akerl>So it's listening. Does your php script load properly through nginx?
19:03<swaj>so now you need to check your nginx.conf and make sure it's passing along to the right port
19:03<jmartinez>is that closing bracket supposed to be there?
19:04<swaj>yeah that's fine
19:04-!-tunabananas [] has quit [Ping timeout: 480 seconds]
19:05<swaj>look in your nginx config for a fastcgi_pass line
19:06<jmartinez>I dont see anything about fast-cgi
19:07<swaj>can you pastie your nginx.conf?
19:07<akerl>jmartinez: Did you actually configure your site to *use* php?
19:07<jmartinez>Here's the current nginx.cof:
19:07<akerl>jmartinez: pastebin your actual site config...
19:07<swaj>yeah we need the file in sites-enabled
19:07-!-oojacoboo [] has joined #linode
19:07<swaj>for your site
19:07<jmartinez>It was working fimne until I removed php5-cgi
19:08<oojacoboo>hey, how can I see what's making our HDD run so hard?
19:08<akerl>That's because it was using php5-cgi, not phpfpm
19:08<swaj>go to /etc/nginx/sites-enabled and pastie the site's config file
19:08<akerl>oojacoboo: iotop
19:08<oojacoboo>ugh, gotta install that :/
19:09<oojacoboo>mmm, can't find that for centos with the repos
19:09<akerl>oojacoboo: Your real problem is centos
19:10<akerl>jmartinez: Your config is full of fail
19:10<oojacoboo>akerl: entirely unnecessary conversation
19:10<bob2>centos :-(
19:10-!-JediMaster [] has quit [Ping timeout: 480 seconds]
19:10<swaj>jmartinez: is this a wordpress site?
19:10<akerl>Very relevant
19:11<jmartinez>yes (wordpress)
19:11<akerl>"IOTop uses features that are a part of a newer kernel build than what CentOS ships with"
19:11<bob2>somewhat mitigated by linode
19:11<oojacoboo>well then
19:11<swaj>jmartinez: I'll make you a config for your site, 1 sec.
19:11<bob2>dunno if centos 5 runs under 2.6.39 though
19:11<jmartinez>Awesome. Thanks
19:11<akerl>jmartinez: For future reference,
19:12<swaj>jmartinez: what's your domain name?
19:12<akerl>Basically, all those if rewrite junks are easily replaced by one tryfiles
19:12<swaj>jmartinez: for the site that is
19:12<tonyyarusso>"ERROR: You appear to be trying to run CentOS in the 21st century."
19:12<swaj>tonyyarusso: rofl
19:12<bob2>problem is that one guy wrote an article in 2005 about wordpress under nginx
19:12<akerl>tonyyarusso: But centos 6 will ship any day now :p
19:12<chesty>including linode
19:12<oojacoboo>bunch of ubuntu kids
19:12<bob2>akerl: lololololol
19:12<bob2>oojacoboo: no
19:12<tonyyarusso>akerl: Of course! Could be tomorrow, I'm sure! :P
19:13-!-Hellojere [~Hellojere@] has quit [Remote host closed the connection]
19:13<bob2>also, centos is objectively a crappy choice for most cases
19:13<akerl>chesty: Does linode's nginx guide use the if nonsense for nginx?
19:13<oojacoboo>but, I use ubuntu on my pc at home
19:13<bob2>a) barely mtaintained
19:13<oojacoboo>it's so easy to use
19:13<bob2>b)_ really old
19:13<bob2>c) really really old
19:13<bob2>d) no commercial backing and small community
19:13<oojacoboo>bob2: you use ubuntu right?
19:13<bob2>oojacoboo: no
19:13<oojacoboo>sci linux/.
19:14<bob2>rhel I can see as a sensible choice if you like that sort of thing
19:14<bob2>oojacoboo: no
19:14<akerl>oojacoboo: Don't assume what distro someone uses just because they mock yours.
19:14<bob2>since at least you get support
19:14<bob2>and it's /less/ out of date than centos
19:14<jmartinez>swaj: did you get that domain name?
19:14<swaj>yeah, check your messages :P
19:15-!-mrpatg [~mrpatg@] has joined #linode
19:15<mrpatg>I've messed up
19:16<mrpatg>Setup new linode with LAMP stack script, im trying to configure apache for a new (first) virtual host, and im getting an error. I am following the tutorial btw
19:16<akerl>mrpatg: Stack scripts are generally not recommended unless you already understand what they do
19:16<akerl>But, what is the error
19:17<mrpatg>Ive setup LAMP before from scratch via command line if thats what you mean
19:17<mrpatg>This is my 2nd or 3rd time ever of doing it
19:18<mrpatg>The error is when i try to reload apache it is saying my site doesnt exist
19:18<akerl>mrpatg: Exact error?
19:18<mrpatg>that is the error, let me give the back info and it will make sense what ive done
19:19<akerl>... Exact error message. Not "it is saying my site doesn't exist"
19:19<akerl>Pastebin the command you run, and the output.
19:20<mrpatg>setting up m VH file, initially put and the alias to non www domain. I didnt reload apache with this info, and following the tutorial, went back and did it to non www first and aliased the www address.
19:20<mrpatg>I removed the www address file, and saved the new non www address file in sites-available
19:20<mrpatg>i am able to a2ensite the site correctly
19:20<priyesh>with iptables how do i allow all localhost traffic (lo) if the INPUT policy is DROP
19:20<akerl>mrpatg: I will not offer any advice until you pastebin the requested info.
19:22<swaj>priyesh: iptables -A INPUT -i lo -j ACCEPT
19:22<priyesh>swaj: do i need a -p all?
19:22<akerl>mrpatg: Does that file exist?
19:22<swaj>priyesh: no, you shouldn't. That should be all you need
19:22<akerl>Then that's why
19:23<mrpatg>I dont know why it is still looking for it
19:23<priyesh>swaj: is the default -p all?
19:23<swaj>priyesh: yes. If you don't specify a protocol, then it uses all
19:23<mrpatg>unless i am mistaken, that file only exists as an alias in my virtual host file
19:24<akerl>mrpatg: `ls -la /etc/apache2/sites-enabled/` and `ls -la /etc/apache2/sites-available/`
19:24<priyesh>swaj: thanks :)
19:24<akerl>And pastebin those
19:24<swaj>priyesh: np
19:26<priyesh>swaj: on a #sudo iptables -L, how do i tell the interface the rule is 'working' on
19:27<jmartinez>Alright, I think I am going to start with a clean slate. What is the best buide to use for getting nginx + php-fpm + wordress?
19:27<akerl>mrpatg: Notice anything?
19:27<swaj>priyesh: not sure I understand what you mean?
19:27<jmartinez>swaj: thanks for all of your help!
19:27<swaj>priyesh: you just want to make sure the rule is working?
19:27<swaj>jmartinez: are you up and running?
19:27<jmartinez>Best guide*
19:28<jmartinez>No still getting 502 error
19:28<swaj>jmartinez: check your php-fpm.log
19:28<priyesh>swaj: yeah.. and also see a visual confirmation
19:28<jmartinez>about to give up and start over from the bottom
19:28<swaj>priyesh: try to ssh localhost and see if it times out :P
19:28<akerl>My bet is that the filename is missing or has an extra slash
19:28-!-shirro [] has joined #linode
19:28<akerl>jmartinez: Pastebin your php-fpm.log, and also your site config
19:28<swaj>akerl: he's got his nginx set up right... his error logs are showing that php-fpm is resetting his connection
19:28<akerl>in nginx
19:29<swaj>akerl: so I think php-fpm is having some issue, but the log file will show it
19:29-!-priyesh [] has quit [Quit: Reconnecting]
19:29-!-priyesh [] has joined #linode
19:29<mrpatg>akerl: thanks
19:30<priyesh>swaj: yeah.. and also see a visual confirmation
19:30<akerl>You're welcome
19:30<swaj>priyesh: try to ssh localhost and see if it times out :P
19:30-!-devcomp [] has joined #linode
19:31<swaj>man php-fpm looks fine from that log
19:31<priyesh>swaj: :)
19:31<priyesh>one last question, when i do $sudo iptables-save > /etc/iptables
19:31<priyesh>it says permission denied
19:32<swaj>because you're piping the output into a file you can't write to
19:32<swaj>the first command is running as sudo, but the second is not
19:32<akerl>jmartinez: Have you restarted nginx since adding the new config?
19:32<jmartinez>I have been messing with this whole systems config all day. Would'nt be surpised if I bumped somthing I shouldn't have earlier. I will just rebuild.
19:32-!-niftylettuce [] has quit [Ping timeout: 480 seconds]
19:32<priyesh>swaj: thought so.. how do i get around that/
19:32<jmartinez>Aker1: yes
19:32<swaj>sudo su -c "iptables-save > /etc/iptables.conf"
19:33-!-frank_usrs [] has quit [Ping timeout: 480 seconds]
19:33<jmartinez>Does anyone have a reccomneded guide for nginx + php-fpm +wordpress?
19:33<priyesh>swaj: thanks again :)
19:33<swaj>priyesh: np :)
19:33<jmartinez>Apparentlt the guide I was using is outdated. :)
19:34<akerl>jmartinez: What distro?
19:34<swaj>jmartinez: what's your host OS?
19:34<swaj>which version?
19:34<mrpatg>jmartinez: i went through this just the other day, getting permalinks to work with wordpress and nginx. Ill see if i can find the link of the guide i used.
19:34<jmartinez>currently 10.4
19:34<swaj>jmartinez: save that config I gave you. I'm certain it will work
19:34<akerl>Because, for distros with up to date nginx, you just install nginx and phpfpm, and pop the fastcgi pass config into your server config
19:35<swaj>jmartinez: with wordpress
19:35<swaj>jmartinez: I know because I use it on my blog :P
19:35<jmartinez>i am just trying to build the3 fastest linode possible for use with wordpress
19:35<jmartinez>Okay, swaj. Thanks
19:35-!-LittPi [] has joined #linode
19:35-!-oojacoboo [] has quit [Ping timeout: 480 seconds]
19:35-!-LittPi [] has quit []
19:35<swaj>so honestly all you need to do is install nginx from the ubuntu PPA. Instal php5-fpm from the ubuntu PPA, and then put in that config file I gave you
19:36<swaj>it should just work (tm)
19:36<akerl>Ugh @ ppas for servers
19:36<swaj>unfortunately the distro repos are out of date horribly, especially with nginx.
19:36<akerl>That's what source is for
19:36<swaj>I compile from source, but PPA is much easier for most people
19:36<jmartinez>how do I install from the ubuntu ppa?
19:36<swaj>I compile PHP and nginx from source myself :P
19:36-!-oojacoboo [] has joined #linode
19:37<akerl>If you're going to use a version outside of the distro repo, you need to ensure you're getting security updates anyway, so you might as well just use source.
19:37<swaj>akerl: the nginx PPA is run by the nginx developer
19:37<swaj>pretty sure that's a safe place to get it
19:37-!-djg320 [] has quit [Ping timeout: 480 seconds]
19:37<jmartinez>how's this guide?
19:38<swaj>jmartinez: -- follow the Ubuntu PPA instructions to get nginx installed
19:38<akerl>jmartinez: You need to stop thinking about this as "create a super fast wordpress/nginx/php server". Build a stable server, add stable nginx and phpfpm, then add wordpress.
19:39-!-LittPi [] has joined #linode
19:39<jmartinez>got it
19:39<akerl>You have to lay the foundation before you start assembling the walls.
19:39<retro|blah>Let's see
19:39<retro|blah>2 days left till the end of the month
19:39<retro|blah>Crapload of quota left
19:39<retro|blah>I think this calls for seeding some Debian isos!!111
19:40<akerl>retro|blah: Or run a tor exit node?
19:40<bob2>akerl: diaf
19:40<retro|blah>I'm not that suicidal
19:40<swaj>jmartinez: here's my best recommendation to you, if you really want to start over. Recreate your linode with 10.04 LTS. Then install nginx from the PPA with those instructions I sent you. Install MySQL from the linode library instructions. And install php5-fpm from the default ubuntu repos. Then just use that config file I gave you and you should be golden.
19:41<akerl>retro|blah: What do you use for seeding from your node?
19:41<jmartinez>swaj: i'm on it! Will report back in a few.
19:41-!-lunks [~lunks@] has joined #linode
19:44-!-Ponch [] has joined #linode
19:46-!-Ponch [] has quit []
19:47<navi>I use uTorrent
19:47<akerl>navi: On a server?
19:47-!-Cromulent [] has quit [Remote host closed the connection]
19:47<navi>uTorrent Linux
19:47<navi>I found it simple enough to run/set up and since I only use it for a bit at the end of a month, it really doesn't bother me
19:48<swaj>I think I might start using my windows server to tunnel crap.... it has 10 TB bandwidth, lol... instead of the 200GB for my linode :P
19:48-!-DephNet[Paul] [~Paul@] has quit [Quit: Leaving]
19:49-!-Rezt [] has joined #linode
19:49-!-Tobarja [~chatzilla@] has quit [Quit: ChatZilla [XULRunner]]
19:49<mrpatg>How would i setup apache so that is my default domain, and goes to, but additional domains go to
19:50<bob2>depends what go means
19:50<devilspgd>I usually do about 3GB/mo on my Linode and I'm debating picking up a second now... Thinking a Tor exit node might be a friendly way to give back to the community
19:50<tonyyarusso>mrpatg: ServerName and DocumentRoot, mostly.
19:50<navi>devilspgd: I prefer torrent seeding and NTP time serving
19:50<akerl>devilspgd: Been there, made that joke, was shot down :p
19:50-!-niftylettuce [] has joined #linode
19:50<akerl>mrpatg: You need to better describe your scenario
19:51<akerl>One IP, or two?
19:51<mrpatg>A single IP
19:51<swaj>man I almost fell for that troll devilspgd :P
19:51<akerl>Then forget all those craps.
19:51<navi>qnza vg
19:51-!-kethry [] has quit [Quit: shell terminated ]
19:51<oojacoboo>akerl: dstat
19:52<akerl>So you want to server one thing, and other domains to all server something else entirely?
19:52<devilspgd>bleh. I just don't troll enough anymore, I'm out of practice.
19:52<akerl>tonyyarusso: yup/
19:52<bob2>devilspgd: it was most excellent
19:53<swaj>devilspgd: you almost had me. I was mid-sentence when I realized that it was a troll :P
19:53<mrpatg>Well just how to setup multiple virtual hosts i guess
19:53-!-priyesh [] has quit [Quit: Reconnecting]
19:53<mrpatg>I have one setup now, but going to my IP and going to that domain, both are returning the same index.html file
19:54-!-priyesh [] has joined #linode
19:54<jmartinez>swaj: apparently I need to install python-software-properties?
19:54<bob2>mrpatg: yeah
19:54<akerl>mrpatg: I still have no idea what you want your end result to be.
19:54<swaj>jmartinez: what's requiring that?
19:55-!-stafamus [~stafamus@] has quit [Ping timeout: 480 seconds]
19:55-!-laser` [] has quit [Remote host closed the connection]
19:55<jmartinez>"If you get an error about add-apt-repository not existing, you will want to install python-software-properties."
19:55<swaj>oh yeah
19:55-!-copperx [] has quit [Quit: Leaving.]
19:55<swaj>because you're adding the PPA repo
19:55<mrpatg>akerl: it will be a challenge wording the questions right. Im trying to ween off of shared hosting and get multiple sites setup on this linode
19:56<swaj>jmartinez: so just "sudo apt-get install python-software-properties"
19:56<oojacoboo>what's a high IO Rate to be concerned about?
19:57<akerl>oojacoboo: High enough that it hurts performance?
19:57<oojacoboo>on a small development server without much public httpd
19:57-!-BP{k} [] has quit [Quit: leavingexit]
19:57<oojacoboo>high enough to be concerned about something funky going on
19:57-!-marcopkb [] has quit [Remote host closed the connection]
19:58<akerl>oojacoboo: It entirely depends on your situation.
19:58<bob2>I worry when IO-wait becomes noticable
19:58<akerl>Higher than the upper bound of normal.
19:58<oojacoboo>well, obv there is a spike, or I wouldn't be addressing it
19:58<oojacoboo>but I can't pinpoint anything
19:58-!-Kuboing [~Administr@] has quit [Read error: Connection reset by peer]
19:58<oojacoboo>swap is clean as nuns pussy
19:59<oojacoboo>logs look ok
19:59<oojacoboo>top isn't running hard
19:59<bob2>what does iotop show?
19:59-!-xt3mp0r__ [~xt3mp0r@] has joined #linode
19:59<oojacoboo>always trollin' huh/
19:59<swaj>silly centos user :P
19:59<bob2>I did mean that seriously
19:59-!-xt3mp0r_1 [~xt3mp0r@] has joined #linode
19:59<bob2>it's basically the thing you need here
20:00<akerl>bob2: Centos doesnt like iotop :p
20:00<swaj>he can't check iotop, because his kernel = teh old :P
20:00<bob2>so get the source and reboot in to a slightly more modern kernel
20:00<oojacoboo>yea, well, I watched dstat and it looked clean as well
20:00<bob2>swaj: changing kernel = select from dropdown in lpm
20:00-!-Kuboing [~Administr@] has joined #linode
20:00<swaj>if he's on paravirt
20:00<akerl>oojacoboo: Remember when I said that centos was your real problem. Still wasn't kidding :p
20:01<hawk>akerl: But it's so enterprisey
20:01<oojacoboo>akerl: well, I am not a sysadmin and certainly have better things to do than fix something that isn't broke just b/c
20:01<oojacoboo>centos is not the issue
20:01-!-xt3mp0r [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
20:01<oojacoboo>it's quite stable if you ask me, I dont' have to worry about all the issues that you guys test out for me
20:01<akerl>oojacoboo: If you're not a sysadmin, you'd better fetch one, before you bork your node trying to play with kernels and source.
20:01*navi is on akerl's side.
20:02*swaj is too :P
20:02-!-xt3mp0r_ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
20:02<oojacoboo>akerl: who says I am playing with kernels or source?
20:02<bob2>get iotop
20:02<akerl>oojacoboo: If you plan on using iotop
20:02<bob2>time is a-wasting
20:02<oojacoboo>akerl: I don't
20:02<swaj>then your I/O problem must not be too bad
20:03<oojacoboo>swaj: 1k
20:03<swaj>ouch :P
20:03<bob2>he's using centos, no need troll and make his suffering worse
20:03<oojacoboo>damn I love this chan
20:04<bob2>anyway, install iotop
20:04<akerl>oojacoboo: Basically, iotop is the tool you need to use to diagnose this problem. The fact that you don't have it is blocking us helping you.
20:04<oojacoboo>snobby ass ubuntu kids
20:04<bob2>trying that will take less time than basically anything else
20:04-!-jake [] has quit [Ping timeout: 480 seconds]
20:04<bob2>oojacoboo: stop it
20:04<navi>oojacoboo: Debian, actually
20:04<bob2>oojacoboo: also stop assuming people who hate centos use ubuntu
20:04<bob2>I hate centos because I /use/ centos
20:04*akerl looks around for ubuntu users.
20:04<navi>I think I see one over there! Kill it!
20:04<oojacoboo>akerl: they are the majority
20:04<akerl>oojacoboo: Not here, but whatever.
20:04<bob2>your imagined stats aside, install iotop
20:05<bob2>then you can see wtf is going on
20:05<oojacoboo>last time I checked the linode stats, that was true
20:05<oojacoboo>by a long shot
20:05<akerl>here = irc channel.
20:05<oojacoboo>Debian is nice
20:05<bob2>amongst linode customers
20:05<bob2>!= irc channel
20:05<bob2>so stop trolling and start installing
20:05-!-frank_usrs [] has joined #linode
20:05<oojacoboo>no can do, not enough time/resources for that atm
20:05<oojacoboo>anyway, thanks for the help
20:06<bob2>well, iotop is the thing to use
20:06<bob2>so if you have time to try to debug it later, get it
20:06<oojacoboo>well, I have to pretend I am living in 2005 and solve the issue like people did then
20:06<oojacoboo>I'll use the wayback machine, thanks anyway
20:06<oojacoboo>so be it
20:06<bob2>pretty sure getting iotop going will be quicker than that
20:07<bob2>not sure why you're so opposed?
20:07<oojacoboo>I have to install a new OS?
20:07<ajmitch>because it's fun to sit on irc & bitch about it?
20:07<bob2>please stop being a fuckwit
20:07<akerl>oojacoboo: No, you need to switch to a newer kernel, and compile from source.
20:07<bob2>get the iotop source
20:07<bob2>making up stuff is very frustrating
20:07<bob2>you're on centos, compiling is something you have to do
20:08<oojacoboo>yea, it's something I try to avoid as well, I am a designer
20:08<oojacoboo>not a sysadmin
20:08<oojacoboo>anyway, I'll monitor it the old way
20:08<swaj>I'm not on ubuntu either, btw
20:08<oojacoboo>thanks for the help
20:08<swaj>I use debian kthx :P
20:08-!-jake [] has joined #linode
20:08<akerl>oojacoboo: Why is the designer doing debug work on the server?
20:08<hawk>oojacoboo: Is there a sysadmin?
20:08<bob2>akerl: how do you think comments like that help
20:08-!-jake [] has quit [Read error: Connection reset by peer]
20:08<oojacoboo>b/c that's what you have to do when you are a small company
20:09<akerl>bob2: Because the sysadmin is the one who needs to be hunted down, so as to install iotop and make progress
20:09<oojacoboo>b/c I am the "sysadmin"
20:09*akerl beats head on table.
20:10<navi>akerl: I'm a physicist as well as a sysadmin. Is that bad too?
20:10<bob2>navi: just don't write code
20:10<bob2>physicist code is only just above engineer code
20:10<navi>bob2: I'm also a mathematician.
20:11<navi>bob2: That makes it worse, right?
20:11<hawk>navi: The horror
20:11<akerl>navi: Only if you're only a sysadmin when it doesn't involve compiling.
20:11<bob2>navi: oh, all your code is in Coq then
20:11<navi>I love compiling
20:11<bob2>navi: so it's useless but provably correct
20:11<navi>Someone else has done the hard bit, I just have to work out a few little teething problems
20:12<akerl>Also, you get to sword fight on wheely chairs
20:12<navi>No, I like watching the progress
20:12-!-bigjocker [~ngranek@] has joined #linode
20:12<navi>Or, as a mathmatician, trying to interpolate the progress from the place in the alphabet it is
20:12<bob2>navi: surely you need to build a statistical model to take into account the letter distribution
20:13<navi>bob2: You don't have that memorised?
20:13<akerl>I imagine there are several whiteboards involved in this process
20:13<bob2>that's all I can remember
20:13<bob2>it's enough to beat "hanging with friends" most of the time
20:13<akerl>Is that for letters in english, or letters in code?
20:13<bob2>akerl: you win
20:14<mrpatg>any suggestions for web based control panels for linode to quickly add domains, emails, etc?
20:14<hawk>I suppose it would be initial letters in source code filenames and/or directories?
20:14<akerl>mrpatg: The invisible, uninstalled kind? command line ftw
20:14<bob2>let me propose a thing
20:14<akerl>hawk: That's what I was thinking. In the same fashion that repos tend to focus on "l"
20:14<bob2>a web control panel that only fiddles a db
20:15<mrpatg>akerl: As much as i enjoy command lining every single thing i want to do, i want my focus on my sites, deploying them, and developing them
20:15<bob2>along with a set up where postfix/dovecot use the db
20:15<bob2>and the apache vhost config and dirs are regenerated from it
20:17<navi>How often does the Linode Control Panel update my bandwidth usage?
20:17-!-jake [] has joined #linode
20:17-!-jake [] has quit [Read error: Connection reset by peer]
20:17<@caker>it's cached for a few minutes - that's it
20:17<akerl>bob2: Something where uninstalling the panel or otherwise ceasing use of it would remove the db, but you'd still have all the original files?
20:18-!-jake [] has joined #linode
20:18<navi>caker: Awesome.
20:19<jmartinez>swaj: Trying to install php-fpm
20:19-!-jake [] has quit [Read error: Connection reset by peer]
20:19<jmartinez>What commands should I use?
20:19<akerl>apt-get install php-fpm?
20:19<akerl>Or php5-fpm, not sure for ubuntu
20:19<swaj>jmartinez: sudo apt-get install php5-fpm
20:20-!-byronb [] has quit [Quit: Leaving...]
20:20<jmartinez>Reading package lists... Done
20:20<LittPi>Need to out, see you all @@
20:20<navi>Reading package lists... Bored. Don't you have any Dickens here?
20:21<bob2>project gutenberg
20:21<swaj>jmartinez: I think the php5-fpm package is actually not in the ubuntu repo for lucid. You'll probably need the brianmercer ppa
20:21<swaj>jmartinez: do this -> sudo add-apt-repository ppa:brianmercer/php
20:22<swaj>jmartinez: then -> sudo apt-get update
20:22<swaj>jmartinez then -> sudo apt-get install php5-fpm
20:22<akerl>swaj: Or just apt-get install php-fpm?
20:22<jmartinez>Going to follow that guide... Yes?
20:23<swaj>akerl: I don't think there is a php-fpm for lucid
20:23<bob2>akerl: well, not even remove the db, but the panel won't fuck up regular config files
20:23<bob2>maybe use puppet to set up the basics
20:23<akerl>Ugh @ old ubuntu. Natty has it
20:23<swaj>jmartinez: that's a good one. follow that guide.
20:23-!-hfb [] has quit [Quit: Leaving]
20:23<hawk>swaj: How advisable is it really to install those things, are they even maintained?
20:24-!-voker57__ [~voker57@] has quit [Ping timeout: 480 seconds]
20:24<swaj>hawk: the brianmercer php PPA is very well maintained and has a good reputation. And the Nginx PPA is maintained by the actual nginx development team, so I'd call that good enough :)
20:24<akerl>bob2: Yea, that's what I meant. So that basically, you have a functional, complete server, minus specifics, on the disk, and the db holds all the configs that are site-specific. mail addresses, vhost configs, etc.
20:24-!-skegeek [] has quit [Read error: Connection reset by peer]
20:25<akerl>So the server can read from the server and apply the settings, without doing any magic on the files that screws you over later.
20:25-!-frank_usrs [] has quit [Quit: ]
20:25<akerl>You'd think someone would have done that already.
20:25-!-skegeek [] has joined #linode
20:25-!-LittPi [] has quit [Quit: CGI:IRC (Ping timeout)]
20:25<hawk>swaj: brian mercer apparently recommends using something else
20:26<swaj>hawk: hmm that's news to me. you can see his revision history
20:27-!-Rezt [] has left #linode []
20:27<hawk>swaj: THat's the very page where I read that.
20:27-!-jake [] has joined #linode
20:27-!-jake [] has quit [Read error: Connection reset by peer]
20:27-!-frank_usrs [] has joined #linode
20:28<jmartinez>swaj: are the recommended php-fpm tweaks ok for a 512 linode?
20:28<hawk>swaj: (The "NOTE: ..." section at the top)
20:28<swaj>hawk: oh man, touche.
20:28<swaj>jmartinez: you should use a different PPA than what that article recommends
20:28<mrpatg>If i want to setup phpmyadmin to be accessed from my site's ip at a particular port, what do i fill in the virtual host file for server name/alias?
20:28-!-Bdragon [] has quit [Quit: Leaving]
20:29<swaj>jmartinez: did you install php5-fpm yet?
20:29<swaj>jmartinez: I recommend then, that you remove it. and put the other PPA in. it's easy to change.
20:30-!-jake [] has joined #linode
20:30<swaj>jmartinez: sudo apt-get purge php5-fpm
20:30<swaj>jmartinez: sudo remove-apt-repository brianmercer/php
20:30-!-jake [] has quit [Read error: Connection reset by peer]
20:31<swaj>jmartinez: those 2 commands will get rid of php5-fpm and the old PPA
20:31<jmartinez>sudo: remove-apt-repository: command not found
20:32<swaj>sigh, sec
20:32<akerl>mrpatg: Are you trying to serve phpmyadmin on the public IP?
20:33<mrpatg>on my single ip address, yes
20:33<akerl>mrpatg: I'd recommend against it.
20:34-!-lunks [~lunks@] has quit [Quit: lunks]
20:34<akerl>mrpatg: As far as setting it up, you'd add another vhost block in your apache configs, pointing at phpmyadmin's location, configured to use php
20:34<swaj>jmartinez: do you have a ppa-purge command? try -> sudo ppa-purge ppa:brianmercer/php
20:34<akerl>mrpatg: That said, I'd very much recommend having phpmyadmin served to localhost, rather than your public IP
20:35<jmartinez>swaj: nope
20:35<jmartinez>same error
20:35-!-Bdragon [] has joined #linode
20:35<mrpatg>i would very much like to just get it up
20:35<+linbot>THAT'S WHAT SHE SAID!
20:36<akerl>mrpatg: Setting it up on the localhost rather than public is no harder on the server side, it just means using an ssh tunnel to use it.
20:37<swaj>jmartinez: do this: sudo apt-get install ppa-purge
20:37<akerl>It seems like it's worth a bit more time for setup, rather than putting it on the public IP and having your box pwned.
20:37<swaj>jmartinez: and then try and run that command again
20:38-!-newbie [] has joined #linode
20:38<+linbot>New news from forums: CPU Spike in General Discussion <>
20:39-!-jake [] has joined #linode
20:39-!-Cruiser [] has joined #linode
20:39-!-jake [] has quit [Read error: Connection reset by peer]
20:39<+linbot>navi: Is Down -> Check if your website is up or down?
20:39<navi>Oh, good.
20:40<navi>Well, not good, but "It's not just me"
20:41<hawk>Maybe it got oracled
20:41<+linbot>acidchild: Is Up -> Check if your website is up or down?
20:42<acidchild>packet harder!
20:42-!-TIBS02 [lemfpomwe@] has joined #linode
20:42-!-byronb [] has joined #linode
20:43<swaj>heckman: you might be interested to know, I talked to my mom again today about being down. apparently some fiber was cut and they were down while it was repaired. the "hacktivist" claims that they took it down were wrong :P
20:43-!-Bdragon [] has quit [Ping timeout: 480 seconds]
20:43<hawk>navi: might be the closest you'll get at the moment ;)
20:44<@ericoc>maybe the hacktivists cut the fiber!
20:44<acidchild>swaj: isn't anycast?!
20:44<swaj>navi: just use libreoffice :P
20:44<swaj>ericoc: lol
20:45<navi>I'm not using either of them
20:45<swaj>seems a lot of distros are switching from OpenOffice to LibreOffice
20:45<hawk>Reading some of the articles, apparently Oracle will retain the openoffice trademark, despite dropping the codebase on apache
20:46<hawk>So maybe there isn't supposed to be a site... even though that seems kind of harsh. Hopefully it's just down for some other reason
20:46<swaj>the site is hosted on apache, so it crashed :P
20:47-!-Rezt [] has joined #linode
20:47<Nivex>are you sure it wasn't because urmom tried to view it?
20:47<+linbot>navi: Yo mommas so big, she knocked Mir out of orbit (805:0/7) [rmumo]
20:47<jmartinez>swaj:E: Couldn't find package ppa-purge
20:48-!-jake [] has joined #linode
20:48-!-blindwaves [] has quit [Quit: Leaving]
20:48-!-mrpatg [~mrpatg@] has quit [Quit: Leaving]
20:48-!-lunks [~lunks@] has joined #linode
20:48-!-lunks [~lunks@] has quit []
20:50-!-TIBS02 [lemfpomwe@] has quit [Ping timeout: 480 seconds]
20:50-!-lunks [~lunks@] has joined #linode
20:51<jmartinez>Swaj: I just removed the brianmercer ppa from the etc/apt/sources and am restarting my linode
20:51<@heckman>swaj: Someone cut fiber? Holy crap, are they fired?
20:53-!-kenichi [] has quit [Remote host closed the connection]
20:55<jmartinez>Swaj: Forget it. I am rebuilding(yet again). What PPA should I use for php-fpm?
20:56-!-Number7 [] has joined #linode
20:56-!-disinpho [] has quit [Ping timeout: 480 seconds]
20:57<hawk>heckman: Assuming it's a "normal" fiber cut I doubt that... Or did you mean if someone at the mastercard networking end is fired?
20:57*akerl smiles at his arch system. PPA you say? :p
20:58-!-djg320 [] has joined #linode
20:58*newbie coming to thank all who tagged-teamed in wrestling a Linode from 9.04 to 10.04LTS :)
20:59<newbie>hawk: cheers.
20:59<newbie>swaj: cheers
20:59<newbie>heckman: cheers
20:59<@heckman>You can highlight us on all one line. :p
20:59*djg320 cracks open a lager
20:59<newbie>wouldn't be as personal would it? :)
20:59<amitz>you can hilight me too.
21:00<newbie>akerl: cheers
21:00*navi adds the missing 'gh' to amitz
21:00<ajmitch>what, no love for the rest of us? :)
21:00<newbie>amitz: sorry..did I miss that you pitched in? if so..sorry..probably in the midst of my sleepiness you said stuff..if so - cheers :)
21:00<newbie>ajmitch: cheers
21:00*newbie not done yet
21:00<amitz>ar ar!
21:01<bob2>dear linode, you should come to pycon-au and buy beer
21:01<newbie>bob2: cheers
21:01<ajmitch>bob2: pycon-au is mid-august, isn't it?
21:01<bob2>ajmitch: yes
21:01<newbie>ok..maybe I should stop flooding..anyhow all CHEERS :)
21:01<bob2>ajmitch: week before pycon-nz I think
21:01-!-disinpho [] has joined #linode
21:01*djg320 highlight entire room and gets banned
21:01<bob2>newbie: oh, glad it is sorted
21:01<ajmitch>bob2: yeah, I think it was arranged to be like that
21:01<amitz>bob2: just tell me if linode actually comes, it should be a good indicator of things to come ;-)
21:02<newbie>bob2: not as glad as me..having said that..I admire all of your patience :)
21:02*ajmitch is registered for kiwi pycon, probably can't get the time off for one in .au as well
21:03<newbie>tell you what though..a few hiccups along the that funky thing with some InnoBD tables getting rogered enroute with the OS upgrade. Weirdness..
21:03<djg320>Is the L5520 @ 2.27GHz
21:03<djg320> superior in architecture over the L5420 @ 2.50Ghz?
21:04<djg320>beastly machine
21:04<swaj>jmartinez: the nginx ppa actually has php5 too
21:05-!-xt3mp0r [~xt3mp0r@] has joined #linode
21:05<djg320>oh wow
21:05-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
21:05<newbie>Got me thinking. I have never had problems with my InnoDB tables do you guys go about doing fixing that? I followed MySQL's suggestion of doing a dump (mysqldump..not me personally) prior to migration then reimporting
21:05<djg320>if I looked close I would have seen, 8mb of L2
21:05<jmartinez>swaj: alright, so the updated rebuild method is..? :)
21:05<bob2>restoring from backup
21:06<swaj>jmartinez: are you back at a fresh ubuntu 10.04 again? :P
21:06<swaj>okay, so the nginx ppa has both of the packages you want.
21:06-!-xt3mp0r__ [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
21:06<newbie>bob2: I haven't had it happen but what if the DB is live and the problem happens whilst in use..restoring from backup will have you lose data?
21:07-!-xt3mp0r_1 [~xt3mp0r@] has quit [Ping timeout: 480 seconds]
21:08<swaj>jmartinez-> sudo apt-add-repository ppa:nginx/stable
21:08<newbie>bob2: not keen on that. having said that I suppose if it is stuffed you're losing data already. That said..when mysqlupgrade reported there were 3 InnoDB tables requiring upgrading, they seemed to still be retrieving and storing data fine without any fixing. What's up with that?
21:08<hawk>djg320: I believe the answer to your cpu question before is "yes"
21:08<swaj>jmartinez: then -> sudo apt-add-repository ppa:nginx/php5
21:08<bob2>newbie: well yes, a db that crashes tables is shit
21:08<hawk>djg320: (but as bob2 said, ark.i.c should answer that)
21:08<bob2>welcome to mysel land
21:08<swaj>jmartinez: then -> sudo apt-get update
21:09<bob2>newbie: can't comment on #2 without seeing the exact log messages, possibly it said "HEY THESE WERE NOT UPGRADED SO I UPGRADED THEM AWESOME BRO"
21:10<jmartinez>Swag: sudo: apt-add-repository: command not found(AHHH!!)
21:10<swaj>jmartinez: ok then -> sudo apt-get install python-software-properties
21:10<swaj>jmartinez: then run those commands
21:12<swaj>alright, so you want to install just two packages
21:12<swaj>sudo apt-get install php5-fpm nginx
21:12<newbie>bob2: thanks for the one who has lived in that land longer I enter as your humble guest :P Yeh I know you're a MyISAM man :) As for the error message - sorry, can't remember it verbatim but it essentially said I needed to upgrade those 3 tables and suggested "repair table" for the job - thing is, that won't work with InnoDB tables. Additionally, those 3 tables that had the error report weren't the only InnoDB tables so I can't see an obvious
21:13-!-xt3mp0r [~xt3mp0r@] has quit [Read error: Connection reset by peer]
21:13-!-xt3mp0r_ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
21:13<bob2>newbie: oh, I'm not recommending myisam
21:13<bob2>it's terrible too
21:13<newbie>"threw up the errors" as in it was always the same 3 InooDB tables
21:13<bob2>everything is terrible
21:13<swaj>jmartinez: okay now put in that config I gave you into sites-enabled
21:13-!-Cromulent [] has joined #linode
21:13*newbie thinks bob2 is sounding like a DB-Emo )
21:14<swaj>jmartinez: like /etc/nginx/sites-enabled/ or something like that
21:14<bob2>newbie: no, just a terrified admin
21:14-!-xt3mp0r [~xt3mp0r@] has joined #linode
21:14<bob2>newbie: who knows which of my slaves are just missing rows?
21:14<bob2>newbie: who knows when writes will start failing because a table crashed?
21:14<bob2>it's just rubbish
21:14<newbie>heh..doesn't admin translate to terrified in assembly?
21:14-!-xt3mp0r_ [~xt3mp0r@] has joined #linode
21:14<HoopyCat>mysql is pretty horrific, all things considered
21:15-!-karstensrage [] has joined #linode
21:15<bob2>sometimes I have to restart replication because it just silently hangs
21:15<newbie>hoopycat: what's less horrific?
21:15<akerl>Was just about to ask for the better option :p
21:15<bob2>if you just monitor the output of 'show slave status', you'll never know
21:15<bob2>you have to have timed writes happening on the master and check when they arrive on the slave
21:15<HoopyCat>i mean, c'mon, mysql's replication falls over and drops/duplicates rows if you look at it funny
21:16<ajmitch>bob2: you're not making me feel any better about using mysql at work
21:16<newbie>hoopycat: stop giving it sleazy stares then :)
21:16<swaj>or fuck sql altogther and use mongodb or couch :P
21:16<HoopyCat>my workaround: completely destroy ill database slaves
21:16<bob2>I don't even know what to do about the silently-missing-rows issue
21:16<hawk>ajmitch: I don't think that was the goal
21:16<bob2>aside from run mk-table-hash or whatever it is every day
21:16<bob2>and reload fucked slaves
21:16<HoopyCat>newbie: i like postgresql, but it has its own foibles (such as having less application support than mongodb or couch)
21:17<HoopyCat>bob2: ok, perhaps more application support than mongodb and couch, but still
21:18<newbie>well..seems to me then that there isn't an obvious reason why those 3 tables consistently threw errors during the upgrade (they were fine on the source linode) and that the only solution to "fixing" the error is doing a restore (pity there's no "repair table" equivalent for InoDB)
21:18<HoopyCat>s/bob2: //
21:18<bob2>the obvious reason is mysql is flakey as fuck
21:18<HoopyCat>newbie: innodb tables generally shouldn't have the sorts of problems that 'repair table' can delete
21:19<newbie>i guess the most worrying thing is silent errors..those tables were happily retrieving and storing data whilst having those "errors" as far as I can tell
21:19<bob2>I agree emphatically
21:20<bob2>afaict you have to scrape your mysql log to detect things before you get emails from customers about thigns not being saved
21:20<newbie>hoopycat: thanks...was thinking that. That's why the whole "error" thing got me thinking - I never had one with an InnoDB table until the upgrade. If indeed it was an error?
21:20<jmartinez>how do I enable a site in nginx? (already in sites-available)
21:21<bob2>add a symlink in sites-available
21:21<amitz>newbie: a periodic dump and restore and compare db is a good practice if you're paranoid enough.
21:21*newbie thinks mysql log scraping sounds like some bizarre Siberian Indigenous hobby
21:21-!-tempesta [] has quit [Ping timeout: 480 seconds]
21:21<newbie>amitz: what's the most efficient way of comparing the db? dump to sql & doing a straight up text-diff?
21:22<amitz>don't know if modern database needs that or is there a better method.
21:22*swaj sticks to MSSQL since I have licenses for it :P
21:22<hawk>bob2: you mean -enabled?
21:22<ajmitch>amitz: we're still debating mysql's status as a modern database
21:22<bob2>hawk: argh yes
21:22<amitz>ajmitch: hah
21:22<ajmitch>it tends to vary based on the storage engine used
21:23<hawk>ajmitch: Has anyone taking mysql's side so far?
21:23<swaj>InnoDB seems to be the "best"
21:23<ajmitch>hawk: sure, we're not saying that mysql using innodb is totally crap
21:23<amitz>newbie: we dumped to sql, restored to another set, and compared row by row. Didn't really think about the performance.
21:23<newbie>hawk: bob2 seems to have a love-hate relationship with it?
21:23<MarkJ>ajmitch: so you're looking at say MySQL v PostgreSQL v ?
21:23<bob2>just hate
21:23<amitz>just inherited old setup.
21:23<newbie>bob2: oh well? ok..forced to live with it at work?
21:24<swaj>MarkJ: vs "Fuck sql, use couch or mongo"
21:24<hawk>newbie, ajmitch, bob2: I think it might be a bit of Stockholm syndrome
21:24<bob2>newbie: yes
21:24<ajmitch>hawk: defending the choices we've made? :)
21:24<newbie>bob2: in your perfect DB world..who'd you be in bed with then?
21:24-!-tempesta [] has joined #linode
21:24<MarkJ>swaj: that's assuming your projects are suitable for NoSQL and not Relational heavy
21:25<bob2>newbie: for most of our crap, postgres, some bits in redis, some bits on disk with that rsync replacement pushing changes out
21:25-!-advion [] has quit [Quit: Leaving]
21:25<seanh-ansca>over heard in the office "i love it when dns issues resolve themselves!"
21:25<newbie>amitz: what operation do you exactly mean by "compared row by row"?
21:26<MarkJ>IMO one of Amazon SimpleDB's biggest issues lies in that you can't do a mass delete, delete has to be by itemName()
21:26<hawk>ajmitch: I don't know how that enters into it(?)
21:26<MarkJ>so you have to do a select (potentially thousands of records) and do individual deletes (or batches of 20) :p
21:26<HoopyCat>MarkJ: 'course, they charge per operation... :-)
21:26<newbie>bob2: ah postgres then
21:26<amitz>select * on both sides, traverse one record at a time to do comparison.
21:27<newbie>amitz: sounds extremely slow...
21:27<bob2>newbie: partly
21:27<bob2>relational databases aren
21:27<MarkJ>but you need a lot of operations to be feasible, surely it must cost more to process each individually!
21:27<bob2>t the be all and end all
21:27<amitz>mind you we didn't have the so called rowset/cache back then.
21:27<HoopyCat>key-value stores are also not the be-all-end-all
21:28<tjfontaine>sometimes you want integrity yaknow :P
21:28<MarkJ>HoopyCat: yup, I like the idea that it's available to me, but there's no way I'd replace my main data stores with it
21:28<tjfontaine>without writing it in the application layer that is
21:28<MarkJ>if anything I'd use it along side the main database for <stuff>
21:29-!-jared0x90 [] has joined #linode
21:31-!-alohatone [] has joined #linode
21:32-!-epochwolf [] has joined #linode
21:33-!-joshdotsmith [] has quit [Quit: joshdotsmith]
21:36-!-maushu_ [] has joined #linode
21:38-!-JSharp [] has joined #linode
21:38-!-maushu [] has quit [Ping timeout: 480 seconds]
21:38-!-seanh-ansca [] has quit [Ping timeout: 480 seconds]
21:39-!-skegeek [] has quit [Quit: Want to be different? Try HydraIRC -> <-]
21:40-!-Kebn [] has quit [Quit: Computer has gone to sleep.]
21:43-!-Cromulent [] has quit [Remote host closed the connection]
21:47<akerl>Ok, so I'm getting a bunch of errors trying to upgrade perl. It's telling me that a bunch of files exist already. I've previously run updates on perl modules using cpan. What's the proper way to fix the current errors?
21:49-!-epochwolf [] has quit [Quit: Bye!]
21:52-!-Rezt [] has left #linode []
21:53-!-epochwolf [] has joined #linode
21:53<amitz>i assume you didn't use your package manager for cpan before, right?
21:54<akerl>? What do you mean?
21:54<bob2>probably give up and force it and then re-cpan them again
21:54<bob2>protip scribbling on /usr will make your package manager unhappy
21:55<akerl>So what's the recommended method for installing perl modules?
21:55<bob2>I use apt-get
21:55<bob2>and dh-make-perl when needed
22:00<jmartinez>what is the best way to move a mysql database from shared hosting to my new nginx linode?
22:01<akerl>jmartinez: Dump it and then import it
22:01<Number7>akerl: Have you checked out CPAN?
22:02<tonyyarusso>jmartinez: If you google for "Backup and restore wordpress database" the WP Codex site has good instructions for doing so (even if you're not using WordPress).
22:02<akerl>Number7: Did you read what I said?
22:02<Number7>Oh, I just reattched to my screen session so I must have missed the first half. Sorry.
22:02<akerl>The problem I had was directly caused because I used cpan :p
22:03<jmartinez>tonyarusso: thanks!
22:03<Number7>akerl: What distribution?
22:07<akerl>:p Tell that to the guy who just spend forever trying to step ubuntu up to be up to date, or the one who had to jump through hoops to get nginx/phpfpm functional
22:07<Number7>I guess I should have asked, users are allowed to assist users here, yeah?
22:07<akerl>That's the idea
22:07<Number7>See that thread.
22:07<amitz>basically, package manager sometimes don't play well withlanguage package manager
22:07<Number7>Let me know if that is relevant to your issue.
22:09<akerl>It is, in the generic "how to fix problem" sense. I'd hoped (foolishly) that perl would expect me to be using cpan or such, and so the update would function around it
22:09-!-tempesta [] has quit [Ping timeout: 480 seconds]
22:09<Number7>imho, cpan is dumber than a box of rocks and perl installs aren't much better
22:10<akerl>No free beer for you :p
22:10<Number7>That bot is quickly training me to avoid saying lol.
22:11<dominikh>Number7: yeah, real quick
22:11-!-louve [] has quit [Remote host closed the connection]
22:11<auraka>Number7: and the sky is blue
22:12<Number7>auraka: Is that in reference to the bot comment or the cpan comment?
22:12-!-tempesta [] has joined #linode
22:12<auraka>cpan comment
22:12-!-maushu_ [] has quit [Ping timeout: 480 seconds]
22:13<Number7>Ahh, no doubt.
22:13-!-febits [] has quit [Quit: Leaving...]
22:13-!-xt3mp0r_ [~xt3mp0r@] has quit [Read error: Connection reset by peer]
22:13-!-xt3mp0r [~xt3mp0r@] has quit [Read error: Connection reset by peer]
22:13<auraka>I love perl but cpan.....cpan deserves a special place in history up there near Windows ME
22:14-!-disinpho [] has left #linode []
22:14<Number7>Ouch, that might be a little too harsh.
22:16-!-protonchris_ is now known as protonchris
22:17<amitz>you can always say col, chuckling out loud ;-)
22:19<encode>or cq
22:19<encode>chuckling quietly
22:19-!-BP{k} [] has joined #linode
22:20-!-BP{k} [] has quit []
22:21-!-BP{k} [] has joined #linode
22:21-!-paparent [] has quit [Read error: Connection reset by peer]
22:22-!-JSharp [] has quit [Ping timeout: 480 seconds]
22:23-!-atan [] has joined #linode
22:27-!-mathew [] has quit [Quit: Leaving]
22:28-!-hfb [] has joined #linode
22:30-!-devcomp [] has quit [Ping timeout: 480 seconds]
22:32<amitz>encode: but the contradiction renders the statement useless!
22:40-!-pharaun [] has quit [Quit: ZNC -]
22:41-!-bigjocker [~ngranek@] has quit [Remote host closed the connection]
22:42-!-pharaun [] has joined #linode
22:43-!-kethry [] has joined #linode
22:46-!-newbie [] has quit [Quit: CGI:IRC (Ping timeout)]
22:47-!-byronbischoff [~byronb@] has joined #linode
22:47-!-Bdragon [] has joined #linode
22:51<jmartinez>When I try to access my linode's IP address through my browser, it shows one of my enabled-sites. Why is it not showing the default nginx page?
22:52-!-byronb [] has quit [Ping timeout: 480 seconds]
22:52<akerl>jmartinez: Because you have it configured not to?
22:53<jmartinez>This is a totally fresh rebuild
22:53<jmartinez>How would I check the configuration to see why this is happenning?
22:53<akerl>The nginx config file?
22:54<Kyhwana>check your nginx config files?
22:55<@heckman> include /etc/nginx/conf.d/*.conf;
22:55<@heckman> include /etc/nginx/sites-enabled/*;
22:55<akerl>jmartinez: Notice anything that's *not* in that file?
22:56<jmartinez>(Newb here)
22:56<jmartinez>Obviously somthing is missing
22:56<Kyhwana>they'll something to do with a default page/setup?
22:56-!-BP{k} [] has quit [Quit: leaving]
22:56<jmartinez>I need to define what file it's looking for?
22:57<akerl>jmartinez: Along the lines of "anything having to do with a default site" or "anything that handles how nginx responds to requests from your ip"
22:57<dcraig>what's in /etc/nginx/sites-enabled?
22:57<akerl>jmartinez: On a more curious note, why do you want nginx to show the default page?
22:57<@heckman>As well as /etc/nginx/conf.d/
22:58<jmartinez>I am trying to access phpmyadmin
22:58<akerl>A) what does that have to do with the default page
22:58<akerl>B) You really don't want phpmyadmin on your public IP
22:58<jmartinez>And it's just sending me straight to of my enabled-sites)
22:58<@heckman>jmartinez: Because that's how you configured it...
22:58<akerl>C) This is something that should go in a virtualhost (server block) in sites-enabled
22:59<jmartinez>Enabled sites: default,, phpmyadmin
22:59<akerl>jmartinez: My bet is that you want to nuke (or move) the default site, for starters
23:00<dcraig> is not pointing to a linode ip :D
23:00<hawk>No shit
23:00<navi>It might be
23:01-!-Perihelion [] has joined #linode
23:01-!-mode/#linode [+o Perihelion] by ChanServ
23:01<akerl>hawk: You say that, but we had someone here before who put in all their conf files, because they saw it in the Linode Library and didn't want to change it
23:01*heckman slaps Perihelion around a bit with a large cactus
23:01<navi>Put that cactus away!
23:01<@heckman>That's not what she said,
23:01<@Perihelion>He can't cactus like I can.
23:02<navi>Saying cactus reminds me of the "What's your bid" prank calls
23:02*Perihelion slaps navi around a bit with a large cactus
23:03-!-byronbischoff [~byronb@] has quit [Read error: Connection reset by peer]
23:03<hawk>akerl: And it wasn't just yet-another-troll?
23:03<jmartinez>Alright guys, so please explain what my nginx.conf should look like
23:04-!-seanh-ansca [] has joined #linode
23:04<akerl>jmartinez: Firstly, stop touching nginx.conf
23:04<akerl>Confine yourself to sites-enabled.
23:04<jmartinez>I have a few sites defined in vhosts including phpmyadmin
23:04<dcraig>you want some "default" page to appear when you visit the site by IP address?
23:04<jmartinez>I havent touched anything
23:04<jmartinez>this is a box stock install
23:04<akerl>Sites-enabled should include all the sites you want to server.
23:04<jmartinez>No i don't want a default page to appear
23:05<dcraig>you might want to disable the "default" configuration in sites-enabled then
23:05<niftylettuce>any folks here use gedit? or you guys vim'ing
23:05<jmartinez>It just was not appearing like it has in previous installs and I thought it may have somthing to do with not being able to access myphpadmin
23:05<akerl>niftylettuce: On my server? vim all the way
23:06<niftylettuce>I am JS/web dev, not python dev... trying to write a plugin for gedit that does gists
23:06<akerl>jmartinez: Pastebin the phpmyadmin config?
23:06-!-Paull [] has joined #linode
23:06<akerl>The one in sites-enabled
23:06<dcraig>I'm surprised phpmyadmin went in sites-enabled and not in conf.d
23:06<dcraig>but I'm not familiar with nginx
23:06<tjfontaine>there's no conf.d concept in the nginx packaging
23:07<akerl>dcraig: Since it's a site, and not a config, I'm not shocked
23:07<tjfontaine>oh wait huh that conf.d dir is new
23:07-!-devcomp [] has joined #linode
23:07-!-seanh-ansca [] has quit []
23:08<akerl>Why is the server "phpmyadmin"?
23:08<jmartinez>It's not
23:09<jmartinez>I see
23:09<dcraig>guess you'd have to go to http://phpmyadmin/ to see phpmyadmin?
23:09<akerl>jmartinez: I'm thinking you want that server_name to be localhost
23:09<dcraig>but you can soon register your own TLDs...
23:10<akerl>dcraig: for a nominal fee :p
23:10<encode>that would be a good tld
23:11<navi>.dot would be so confusing
23:11<dcraig>heh, there's a howto that suggests adding "phpmyadmin" to your /etc/hosts file
23:11<dcraig>that sounds sloppy
23:11<navi>especially if your website is 'dotdotdot' with subdomain dot you'd have
23:11<Kyhwana>dcraig: for a cool few hundred thousand $
23:12<akerl>dcraig: Not really. I do something similar to serve my admin stuff on localhost
23:12<dcraig>if you want to be able to access phpmyadmin at, then you migth add this to the config file for location /phpmyadmin { alias /usr/share/phpmyadmin/; }
23:12<dcraig>or something
23:13<dcraig>of course, maybe you don't want phpmyadmin to be easy to get to...
23:13<akerl>Meh @ phpmyadmin on public IP. Great way to get your server compromised
23:13<akerl>That's why I said he probably wants it on localhost
23:13<navi>are we allowed numbers in the tld?
23:13<Kyhwana>or .lan
23:13<encode>akerl: same could be said of php* on public ip
23:13<dcraig>but then he has to figure out how to access it on localhost
23:14<akerl>encode: Meh. phpmyadmin is a known target.
23:14<dcraig>might be easier to just restrict access to phpmyadmin by IP or something
23:15<dcraig>or just use /secretphpmyadmin instead of /phpmyadmin
23:15<dcraig>that'll fool 'em
23:15<navi>shorten it to pma
23:15<navi>nobody will guess
23:15<encode>they will now
23:15-!-Rezt [] has joined #linode
23:15<encode>now i have to go change all my urls
23:16<akerl>Or (just rambling here) why not use a desktop admin tool, that already supports connecting via tunnel, and solve all the problems at once.
23:16-!-CyberCholo [~androirc@] has joined #linode
23:16<encode>or just learn to use sql properly
23:16-!-CyberCholo [~androirc@] has quit []
23:17<dcraig>jmartinez wants to solve the problem of getting to phpmyadmin via, and y'all are just making more problems for him :p
23:17<@heckman>Woah, is the Google bar black for any one else?
23:17<akerl>heckman: yea
23:17<dcraig>black for me
23:17<encode>heckman: yes
23:17<dcraig>it's snazzy!
23:18<@heckman>I don't really like it...
23:18<ajmitch>heckman: google are turning evil
23:18<@heckman>It really looks stupid with the white background...
23:18<encode>i don't really care
23:18<dcraig>maybe there's a greasemonkey script to switch it back
23:18<akerl>I'm a fan of dark bg, light text. Wish they'd do it with the whole page.
23:18<dcraig>remember when they forced background images on everyone?
23:19<encode>give it a few years, and it will start looking like altavista
23:19<dcraig>I like that the top bar is now more consistent across all their services
23:19<mtkoan>is it just me or is borked right now
23:19<akerl>dcraig: Except
23:19<+linbot>akerl: Is Up -> Check if your website is up or down?
23:19-!-BP{k} [] has joined #linode
23:19<mtkoan>i didn't say down, just borked
23:19<akerl>mtkoan: borked how?
23:19<mtkoan>certain pages, 500ing
23:20<dcraig>what's a page I should click on to see a 500 error?
23:20<mtkoan>search just print php parse errors
23:20<encode>which certain pages?
23:20<Rezt>I've just borked my static ip config and am trying to repair it from the ajax console, how can I get out of vim insert mode without the escape key in the ajax console? as it isn't being captured
23:21<encode>mtkoan: try the au2 version
23:21<encode>it works :)
23:21<mtkoan>gmmmm it appears to be us2 subdomain, us3 works too
23:21<encode>gg australia
23:21<mtkoan>fail us!! :)
23:21<dcraig>why no escape key?
23:21<akerl>Rezt: The better question is "why no escape key"?
23:21-!-byronb [] has joined #linode
23:21<+linbot>New news from forums: Kernel Panic triggered by VPN login in General Discussion <>
23:21<Rezt>ajax console doesnt seem to recognize it
23:21<dcraig>try Ctrl+[
23:21<Rezt>that and backspace
23:22<retro|blah>Rezt: What browser
23:22<retro|blah>Try with firefox
23:22<@heckman>use Mozilla Firefox
23:22<dcraig>so did that work?
23:22<@heckman>Google Chrome doesn't render Ajaxterm properly.
23:22<akerl>Rezt: Both escape and backspace work for me in chrome ajaxterm
23:22<retro|blah>And lish via ssh is better than via the ajaxterm if you can help it
23:22<hawk>Or just use ssh for accessing lish
23:22<@heckman>akerl: Lish Ajaxterm has *never* worked for me on Chrome.
23:23-!-Aexoden [] has quit [Quit: Goodbye, cruel world!]
23:23<akerl>heckman: Seriously? Lemme check again, maybe I dreamed it
23:23-!-Aexoden [~Aexoden@2600:3c01::f03c:91ff:fe96:478c] has joined #linode
23:23<@heckman>I use Chrome for everything, and I need to keep Firefox on my iMac at work just to use Lish Ajaxterm is needed.
23:23-!-Paull [] has quit [Quit: CGI:IRC (Ping timeout)]
23:24<Rezt>ha, well in safari at least backspace works
23:24<dcraig>did Ctrl+[ work?
23:24<akerl>I like that I get "bad or expired session" if I try to ajaxterm while tunnelled through my node
23:24<Rezt>maybe something to do with keyboard layout am using european mac
23:25<Rezt>no ctrl+[
23:25<hawk>Rezt: No, it's something with ajaxterm and to some degree browser (in)compatibility
23:25<hawk>Rezt: ssh is the most robust way of using lish
23:26<akerl>heckman: Wow. Definitely borked. I swear it used to work
23:26<akerl>I guess I haven't had to ajaxterm since I switched from firefox
23:26-!-rideh [] has quit [Quit: Leaving]
23:27<dcraig>maybe you guys could have some information about browser compatibility for ajaxterm on the site
23:27<navi>Compatible with: "Nothing"
23:27<navi>* Only Nothing versions 2.1 and greater
23:28<@heckman>I've always been curious why Chrome seems to puke on the Ajaxterm.
23:29<Rezt>just did it with lish via ssh in the end.
23:29<hawk>I've always been curious if there is any browser where ajaxterm fully works
23:29<@heckman>Firefox is probably as close as you're going to get
23:29<niftylettuce>does anyone know the easiest way to "stat" a port, i.e. determine if there is a running service?
23:29<akerl>hawk: Netscape?
23:30-!-failure_wav [] has quit [Quit: ZNC -]
23:30<akerl>niftylettuce: netstat?
23:30<alohatone>netstat -ap
23:30<hawk>akerl: Really?
23:30<@heckman>netstat -apln = my fav :p
23:30<akerl>hawk: My grandfather used netscape communicator for his email up until like a year ago.
23:30<Rezt>lsof -i
23:31<@heckman>Er, remove the `l`.
23:31<hawk>heckman: I was just thinking that a and l was a strange combo
23:31<dcraig>think I'll give netscape 4 a try
23:32<@heckman>hawk: 'a' includes 'l'. I am used to not using 'a'.
23:32-!-failure_wav [] has joined #linode
23:35<hawk>heckman: But l excludes part of a... but lets drop this
23:36<@heckman>I thought using 'a' pretty much included what would be shown if 'l' was used on its own.
23:38<hawk>heckman: well.... neither specified = non-listening only, l specified = listening only, a specified = both listening and non-listening
23:39<@heckman>Right right. I was saying I use 'l' often, and not 'a'. So my mind wasn't thinking when I typed it because I am used to using 'l' so much.
23:39*ajmitch is too used to using -nutap
23:39<dcraig>i just put all the letters so I get everything !
23:40<navi>ajmitch: Nut Tap?
23:40<navi>Sounds painful
23:40<dcraig>it's a gentle tap
23:40<ajmitch>navi: :P shows tcp & udp
23:42-!-tempesta [] has quit [Ping timeout: 480 seconds]
23:42-!-tempesta [] has joined #linode
23:44-!-failure_wav [] has quit [Quit: ZNC -]
23:45-!-Cruiser is now known as Cruiser|Afk
23:47-!-andrew [~andrew@] has joined #linode
23:55-!-devcomp [] has quit [Quit: devcomp]
23:55-!-devcomp [] has joined #linode
23:56-!-failure_wav [] has joined #linode
23:59-!-VS_ChanLog [] has left #linode [Rotating Logs]
23:59-!-VS_ChanLog [] has joined #linode
23:59-!-jared0x90 [] has quit []
---Logclosed Wed Jun 29 00:00:35 2011