Back to Home / #linode / 2006 / 05 / Prev Day | Next Day
#linode IRC Logs for 2006-05-29

---Logopened Mon May 29 00:00:21 2006
00:10<fo0bar>The following errors were generated:
00:10<fo0bar># Not a valid domain name.
00:10<fo0bar>it won't let me add
00:10<fo0bar>I was going to point the subdomain to linode with NS records
00:17<fo0bar>aks, I can't get anything to accept for a cname
00:17<fo0bar>err also
00:18<@caker>am I being dense here, or should it accept for a "domain zone" ?
00:18<encode>i think not, but im no dns expert
00:23<fo0bar>caker: yes, I can simply do "mikegrbsucks IN NS" (and 2) on my nameservers to delegate them to you
00:23<@caker>cnames fixed
00:23<fo0bar>also, there are too many exceptions to do filtering by inclusion... they're all "domains"
00:23<@caker>hmm, so that means I need to add another field for NS records, too
00:24<fo0bar>MX too, you can have MX records for subdomains
00:25<@caker>ok, easy enough to add
00:27<@caker>not very intiutive for most users, but the fields will be there..
00:27<fo0bar>also, I just thought of this... how about a link to see what the generated zone file looks like ("export to text" or something)
00:27<@caker>yeah, that'll come after I get the templates generating, and it would be nice to have viewable logfiles from the ns servers, too
00:28<TheFirst>caker: i still say you'd save a lot of trouble with djb ;)
00:28<@caker>TheFirst: honestly, what's the difference? I mean, if I had correctly generated zone/config files for either or, why would one be better?
00:29<fo0bar>TheFirst: umm, does djbdns even have the capability to do AXFR?
00:29<fo0bar>I thought it was all about "why use a standard DNS transfer protocol? use rsync instead!"
00:29<TheFirst>fo0bar: yes
00:29<TheFirst>i do it to sync to zoneedit (my slave)
00:29<TheFirst>axfrdns and axfrget
00:30<TheFirst>caker: it's easier to manage the files via scripts as you would want to in this case
00:30|-|square [] has quit [Quit: ]
00:30<TheFirst>caker: i'm writing code that is doing the same thing you plan to write now....
00:30<@caker>TheFirst: not really managing them, just spewing them out from a template .. find/loop/replace, save to a file...
00:30<@caker>same as I do for my rdns zones
00:31<TheFirst>i suppose....bind is just a bit more coding to write the files
00:31<fo0bar>caker: to avoid incrementing the serial every time a small change is made, if there are any unsaved changes the top of the page should say "you have unsaved changes, click here to apply, etc"
00:32<@caker>fo0bar: yeah, planning on doing that ... having a third state "Changes made, click to activate", besides ACTIVE and DISABLED
00:32<fo0bar>and speaking of which, while creating the zone for the first time, there should be an optional field to specify a starting point for serials, in case the user is transferring from another dns provider
00:33<@caker>why would that matter, exactly?
00:33<TheFirst>if transferring from another dns provider set linode as slave, get the records, then convert to master setup...that'd be the easiest and solve that problem i think
00:33<fo0bar>in general: secondaries don't like it when serials go down between updates
00:35<@caker>fo0bar: so you're saying an external dns server that was master for the zone (and still has that file), and now it's switched to slave, it would still use the "old" master file to compare whether to get updates from the (now master) linode ns servers?
00:35<fo0bar>yes, and will keep it until the expire period (which could be weeks)
00:35<TheFirst>in my experience with zoneedit going master->slave wipes the records
00:36<TheFirst>caker: how often do you plan to get zones in the case of a slave setup?
00:37<@caker>TheFirst: that's set via the SOA refresh record of the zone
00:38<fo0bar>caker: DJB doesn't believe in SOA. it's a tool of the devil
00:38<@caker>fo0bar: how would I successfully handle that situation? I'd rather use the YYYYMMDDnn format, but if they enter a start serial that's insanely high, I'd never be able to revert back to that format
00:39|-|elw00d [] has quit [Ping timeout: 480 seconds]
00:39<fo0bar>caker: hmm, good point... while I like YYYYMMDDnn, it could become a problem with over-zealous users committing... I was thinking N+1 when making that comment
00:41<@caker>that's easily worked-around by only checking if I need to regenerate zones every hour or something
00:42<@caker>or 15 minutes... 96 potential updates in one day
00:43<fo0bar>that could work, but don't just regenerate blindly of course... only if there are changes
00:44|-|ifpg [] has joined #linode
00:45<ifpg>anyone on host59 having problems?
00:45<ifpg>the job queue doesn't appear to be working
00:46|-|IntuiWorx [] has quit [Quit: This computer has gone to sleep]
00:47<@caker>ifpg: one sec
00:52<@caker>ifpg: ok, had to turn off the "disable updatedb" filesystem helper -- grep likes to hang up on special files
00:52<ifpg>thanks chris... I appreciate your time
00:53<@caker>np -- we'll look into fixing that
00:53[~]efudd took a $40 lesson on "how to miss shots" in pool tonight ><
00:54|-|ifpg [] has quit [Quit: CGI:IRC]
00:54<@caker>wtf is a p------- file?
00:54<@caker>prw------- 1 root root 0 2006-05-29 01:24 control
00:54<efudd>oh. pipe.
00:55<@caker>oh, duh
00:55[~]efudd kicks caker in the nadsw
00:55<efudd>with that said, thanks for the pointers on udev! :)
00:55<@caker>hmm, need grep to skip those (with -r)
00:59<efudd>a 47yr old man called me "dusty" tonight for calling the games at $40 lost :/
00:59[~]efudd feels 80
01:01|-|FireSlash [] has quit [Ping timeout: 480 seconds]
01:01<efudd>I almost got in big shit tonight. I take off my wedding ring when I shoot pool (and other times...). Tonight I managed to _lose_ my ring.. found in a few chairs over in the floor. :/
01:20<TheFirst>caker: add a grep -v prior ?
01:21<@caker>-r does recursive directory searches, -v is negative pattern mathing (not on files, on contents)
01:21<@caker>it's when it hits a pipe file is when it hangs, so -v wouldn't help there
01:22<@caker>"-D skip" is supposed to work on FIFOs, but has no effect on those pipe files ... :<
01:24<Battousai>maybe it's one of those newfangled LIFO pipes ;)
01:30<efudd>a pipe has no EOF, therefore, well, /that.
01:36<warewolf>caker- find . -type f -exec grep string {} \;
01:36<warewolf>caker- '-type f' says regular files only
01:42<TheFirst>that the command i was trying to remember
01:43<TheFirst>er chain i guess
01:46<warewolf>find is a very powerful utility
01:46<warewolf>especially gnu find
01:54|-|spr [] has quit [Quit: Spoon!]
02:02<warewolf>by applying the corporate IT security policy to a server in a state one westward of me, I locked myself out of it.
02:02<warewolf>the fucking policy disabled terminal services.
02:07[~]caker bounces host52
02:14<@linbot>New news from forums: Reboot: host52 in System and Network Status <>
02:15|-|cmantito [] has quit [Ping timeout: 480 seconds]
02:16|-|sprouse [] has quit [Ping timeout: 480 seconds]
02:32<@caker>all hail netboot+initrd
02:33|-|coumbes [~coumbes@] has joined #linode
02:37<warewolf>or etherboot?
02:37<warewolf>there's a grub for pxe
02:44|-|Aerv [] has joined #linode
02:44<Aerv>Hi, any info on how long host52 will be down?
02:46<warewolf>caker just rebooted it
02:46<warewolf>he's aware
02:47<Aerv>Cool, thanks.
02:49|-|Dreamer3 [] has quit [Quit: This computer has gone to sleep]
02:50|-|Dreamer3 [] has joined #linode
02:54|-|Aerv [] has quit [Quit: "A motion to adjourn is always in order." - Lazarus Long]
03:24[~]caker moans
03:37|-|paul [] has joined #linode
03:38<paul>Caker.How long should I wait for my linide on host52 to start?
03:40|-|paul [] has quit [Quit: ]
03:40|-|paul [] has joined #linode
03:44|-|paul [] has quit [Quit: ]
03:44|-|LinodeJavaUser [] has joined #linode
03:44<LinodeJavaUser>caker. My linode paul will not start
03:44|-|LinodeJavaUser [] has quit [Quit: ]
03:45|-|paul [] has joined #linode
03:46<@caker>paul: give it a few more minutes
03:49|-|konoko [] has quit [Ping timeout: 480 seconds]
03:50|-|paul [] has quit [Quit: paul]
03:52|-|fake [] has quit [Ping timeout: 480 seconds]
03:54|-|AztekRW [] has joined #linode
03:58<AztekRW>ITS working now.
04:14|-|Makenai [] has joined #linode
04:17|-|konoko [] has joined #linode
04:46|-|cmantito [] has joined #linode
04:48|-|cmantito [] has quit [Quit: ]
04:48|-|cmantito [] has joined #linode
04:53|-|Makenai [] has quit [Quit: CGI:IRC (Ping timeout)]
05:11|-|womble [] has joined #linode
05:14|-|AztekRW [] has quit [Quit: ]
05:37|-|Dreamr3 [] has joined #linode
05:43|-|Dreamer3 [] has quit [Ping timeout: 480 seconds]
06:05|-|jekil [] has joined #linode
06:42|-|Netsplit <-> quits: kriby, Jay-dargo, alnr, besonen, Ciaran, schultmc, Battousai
06:47|-|Netsplit over, joins: Ciaran
06:47|-|Netsplit over, joins: besonen
06:48|-|Netsplit over, joins: schultmc
06:48|-|Netsplit over, joins: Jay-dargo
06:48|-|Netsplit over, joins: alnr
06:48|-|Netsplit over, joins: kriby
06:48|-|Netsplit over, joins: Battousai
07:12|-|Netsplit <-> quits: kriby, Jay-dargo, alnr, besonen, Ciaran, schultmc, Battousai
07:13|-|Netsplit over, joins: Ciaran
07:13|-|Netsplit over, joins: besonen
07:13|-|Netsplit over, joins: schultmc
07:13|-|Netsplit over, joins: Jay-dargo
07:13|-|Netsplit over, joins: alnr
07:13|-|Netsplit over, joins: kriby
07:13|-|Netsplit over, joins: Battousai
07:45|-|coumbes [~coumbes@] has quit [Quit: Leaving]
08:44|-|IntuiWorx [] has joined #linode
08:45|-|Netsplit <-> quits: kriby, alnr, Jay-dargo, besonen, Ciaran, Battousai, schultmc
08:45|-|Netsplit over, joins: kriby
08:45|-|Netsplit over, joins: Jay-dargo
08:45|-|Netsplit over, joins: Ciaran
08:45|-|Netsplit over, joins: Battousai
08:45|-|Netsplit over, joins: schultmc
08:45|-|Netsplit over, joins: alnr
08:46|-|besonen [~besonen@] has joined #linode
08:46|-|darkbeholder [] has quit [Remote host closed the connection]
09:01|-|darkbeholder [] has joined #linode
09:10|-|konoko [] has quit [Ping timeout: 480 seconds]
10:02|-|Kurt [] has joined #linode
10:05<Kurt>is there any way I could mail in a check to pay for a linode rather than use a credit card?
10:16|-|luis_ [~fortytwo@] has quit [Quit: leaving]
10:20<JasonF>caker: mikegrb: ^^^
10:21|-|Kurt [] has left #linode [The problem with the world is not political but philosophical--it's not that the wrong people are doing things, but that the wrong things are being done]
10:55|-|konoko [] has joined #linode
10:57<@linbot>New news from wiki: User talk:Harleypig <>
11:16|-|spr [] has joined #linode
11:38<@linbot>New news from forums: 2.6 kernels that work with the Gentoo distribution? in General Discussion <>
12:10|-|afv-13 [] has joined #linode
12:24|-|NeonNero [] has joined #linode
13:06|-|IntuiWorx [] has quit [Quit: Leaving]
13:16<@caker>Time remaining: 4:53 <-- new tibook battery :)
13:17<NeonNero>sweet :)
13:17[~]taupehat needs a lottery win so he can get a macbook pro
13:17<taupehat>Also, I need to go into work on my day off
13:17<taupehat>the PDC (actually, the ODC) fell over
13:18<NeonNero>so, has anyone read this article yet? ->,aid,125772,00.asp
13:18<NeonNero>"The 25 Worst Tech Products of All Time"
13:18<taupehat>I'd say IIS should have made that list
13:18<taupehat>and any web app that uses commas in links
13:18[~]taupehat grumbles...
13:18<NeonNero>i'm just sitting here and nodding my way through the list
13:19<@caker>taupehat: ODC?
13:19<taupehat>caker: Only Domain Controller
13:20<@caker>in other words..?
13:20<taupehat>the feckin LDAP box is dead
13:20<@caker>windows thing?
13:20<Eman>they list winme but dont list vista?
13:20[~]caker gives you some coffee
13:20<taupehat>I've got a server on order for July that's going to be a backup domain controller
13:20<taupehat>among other things
13:21[~]taupehat drinks the coffee
13:51<JasonF>ldap is fun
13:51<JasonF>kerberos is evil
13:51<JasonF>at least MIT kerb
13:59|-|FireSlash [] has joined #linode
14:06<fo0bar>caker: linux doesn't list the discharge rate in /proc/acpi for my laptop, but from some quick math, I get just over 5 hours
14:07<afv-13>sounds porno
14:09<@caker>fo0bar: nice
14:09<@caker>I'll take 5hrs over 5 minutes
14:09<fo0bar>5 minutes is just about the time for my first laptop, a vaio
14:15|-|FireSlash [] has quit [Quit: Leaving]
14:31|-|spr [] has quit [Quit: Spoon!]
14:53|-|spr [] has joined #linode
15:04|-|Dreamrr_3 [] has joined #linode
15:10|-|Dreamr3 [] has quit [Ping timeout: 480 seconds]
15:27|-|Battousai [] has quit [Ping timeout: 480 seconds]
15:27|-|oneman [] has joined #linode
15:28<oneman>any opers aboot?
15:35<@caker>what's up?
15:39<oneman>I was just wondering if you could tell if my IO usage has been lesser by a bit in the last week
15:39<oneman>since I disabled postgres autovaccume
15:39<oneman>it was claimed I was the 3rd highest IE user
15:39<oneman>IO I mean
15:40<@caker>it looks ok now, you're at 90M swap which might be contributing to load
15:42<oneman>I have 70megs of ram free
15:42<oneman>so whats in swap and why I dont know
15:42<oneman>I dont think its doing anything but sitting there
15:44<oneman>i don't really understnad the postgres autocaumme etc stuff
15:44<oneman>I will switch that site to mysql 5 at some point
15:44<oneman>i do like the postgres command line tools
15:44<@mikegrb>MemFree: 3912 kb
15:44<oneman>they rock
15:44<@mikegrb>where is 70mb from?
15:45<oneman>let me just try cycling the rails site and see what changes
15:46<oneman>well that changed everything
15:47<oneman>I must have a memory leak or something
15:47<oneman>or postgres is somehow causing it
15:47<oneman>i want to blame postgres for everything
15:49<oneman>cycling the rails site and postgres got rid of all swap
15:49<oneman>with even 15megs free on top
15:50<oneman>when I switch to xen ill migrate to mysql if not soon
15:50<oneman>I dont mean to be bad to the host machine
15:51<oneman>im sure there is a good amount of legitamate traffic, but I dont mean for there to be any silly things going on
15:52<afv-13>what did top say was using all the ram?
15:55<oneman>looks like its the two disbatch.fcgi
15:55<oneman>Ok I ll migrate to mongrel and mysql5 and all my problems will be solved ;]
15:55<oneman>lighttpd + mongrel rather than lighttpd + fastcgi
16:03|-|oneman [] has quit [Remote host closed the connection]
16:25|-|NeonNero [] has quit [Quit: Everybody who believes in telekinesis, raise my hand!]
16:33|-|womble [] has quit [Quit: What's behind the round window...]
16:55|-|Spads [] has joined #linode
17:56|-|FireSlash [] has joined #linode
18:08<@linbot>New news from forums: Cannot shutdown or reboot properly via LISH in General Discussion <>
18:24|-|womble [] has joined #linode
18:27|-|jekil [] has quit [Quit: Leaving]
18:38<@linbot>New news from forums: a general Linode and Xen question in Xen Public Beta <>
18:44|-|FireSlash [] has quit [Ping timeout: 480 seconds]
18:51|-|interferon [] has joined #linode
18:56|-|interferon [] has quit [Quit: ]
19:38|-|schultmc_ [] has joined #linode
19:50<@linbot>New news from forums: Reboot: host31 in System and Network Status <>
19:56<@caker>^-- old
19:59|-|schultmc_ [] has quit [Quit: Client exiting]
20:13|-|sprouse [] has joined #linode
20:50|-|Antitribu [] has joined #linode
20:58|-|Antitribu [] has quit [Ping timeout: 480 seconds]
21:01|-|JonR [] has joined #linode
21:02<JonR>has the tokens_max changed, or did I get throttled down?
21:09<JonR>token_refill=100 token_max=20000
21:19<@caker>JonR: throttled
21:20<@caker>JonR: you're > 100M into swap
21:20<JonR>I see.
21:20<@caker>my suggestion: tune apache, mysql etc -- need some help?
21:20<JonR>I've tuned apache based on your conf file.
21:21<JonR>I've tuned mysql based on another conf file I found on the forums.
21:21<@caker>any other memory hungry apps running?
21:22<JonR>I've got BIND running as a dns cache, since I have apache and courier doing reverse DNS. Can't tell if it's helping performance any, though.
21:23<JonR>(for what it's worth, I sure would have appreciated an email or you opening a support ticket when the throttling-down happened; I didn't notice until I tried to do an apt-get update)
21:24<JonR>here's my psaux:
21:29|-|Antitribu [] has joined #linode
21:29<JonR>hm, I have NO idea why some of this is running. :(
21:40<JonR>okay, I've got it down to 41mb swap used.
21:40<JonR>still having problems with my fastcgi being slow, though.
21:46|-|Dreamrr_3 [] has quit [Quit: This computer has gone to sleep]
21:47|-|Dreamrr_3 [] has joined #linode
21:57<TheFirst>JonR: do you need apache, etc? cut down to the lesser resource hog options if not
21:58<JonR>sadly, I need apache and mysql.
21:58<TheFirst>why apache? what does it do that lighttpd wouldnt?
21:58<JonR>because I use wordpress for my blog, which needs those. and typo, which I'd considered switching to, doesn't bloody work.
21:58<TheFirst>i run wordpress on lightty
21:58<JonR>TheFirst: it has actual documentation for mod_rewrite
21:59<JonR>I like lighty. I wish them well, and hope they'll improve their docs to a point where I can use them for more than trivial things.
22:00<JonR>at the moment, though, my attempt at a lighty configuration was half-broken, slower than apache, with little visible memory benefits.
22:01<TheFirst>odd ... saved huge amounts of mem and handled load much better in my setup
22:02<taupehat>it has to also make my coffee for me in the mornings
22:02<JonR>my setup is almost all python fastcgi stuff, with wordpress as the only php app. which I'd love to switch to another blogging system if I could find one with a wordpress importer.
22:03<JonR>anyhow, I killed off some processes left over from when I was trying lighty (like the php fastcgi process) and the BIND cache that wasn't caching, and I'm down to less than 40mb in swap
22:03<TheFirst>btw, what's the problem with the rewrite docs? (not trying to cause an argument...just curious)
22:04<TheFirst>at the risk ofa beating from mikegrb djb over bind will save you space also
22:04<@caker>JonR: thanks, I appreciate it. I'll notify you next time this happens... I've reset your token refill/max, but it looks like you're still burning tokens
22:04<JonR>caker: yeah, I'm running apt-get dist-upgrade at the moment
22:05<JonR>TheFirst: well, from what I'd read, the best way to get wordpress fancy urls on lighty (it uses mod_rewrite on apache) is to use the Power Magnet thingie
22:05<JonR>caker: thanks
22:06<TheFirst>naw only one small mod to make...
22:06<TheFirst>gimme a min, i'll grab the line i use
22:06<JonR>the Power Magnet uses Lua. I couldn't find any docs at ALL for how to write comparable rewrite rules.
22:06<JonR>I asked in IRC, the maintainer said it was obvious, and that apache uses an obscure language for their rewrite stuff; at least he chose a standardized language. needless to say, I found this less than helpful.
22:07<TheFirst>lighttpd rewrites are obvious
22:07<TheFirst>basically near identical to perl code
22:07<JonR>I don't know perl. :P
22:07<TheFirst>well it's variable =~ regex
22:08<JonR>but I learned apache rewrites through loads of examples in the docs.
22:08<JonR>lighty has perhaps 5% of the amount of useful examples as apache has, in my experience
22:08<TheFirst> ... that is pretty comprehensive even if short
22:08<TheFirst>lightty you have to toy with a bit more because you have to do things its way whereas apache molds to your way ...
22:09<JonR>that's another thing that annoys me about it.
22:09<JonR>but I'd be satisfied with replacing mysql with postgres
22:09<TheFirst>well that's a performance/resource saver
22:09<TheFirst>url.rewrite = ( "^/archives/(.+)$" => "/index.php/archives/$1" )
22:09<TheFirst>that's the only line i needed to get word working nice
22:11<JonR>I wonder why I couldn't find that anywhere on the net.
22:11<TheFirst>it's in the wordpress forums
22:12<TheFirst>think a google for "lighttpd wordpress" or somesuch was all i needed
22:12<JonR>also... my urls take the form or
22:12<JonR>TheFirst: did that several times... :(
22:12<TheFirst>yah gimme a sec i'll find the had ways to make it look more like yours
22:13<TheFirst>actually i don't know that the way i have wouldnt work with yours too
22:13<TheFirst>just a bit of a change since you wouldnt have the archives part
22:20<TheFirst>fwiw wordpress can also do perma linkes without mod_rewrite
22:20<TheFirst>which is what that rewrite rule above does..
22:21<JonR>I want to get away from php completely...
22:21<JonR>I hates it.
22:23<TheFirst>but yah python cgi works fine with lightty too...really that'd save you a lot of resources
22:25<TheFirst>one thing that will save...ditch sshd in favor of dropbear...
22:26<JonR>tried that. it didn't like my keys
22:26<TheFirst>misconfigure then...
22:26<JonR>ssh-dss format
22:26<JonR>the docs only ever mentioned rsa, so maybe it doesn't like dss
22:27<TheFirst>i use ssh-dss
22:27<TheFirst>so i know it works ;)
22:27<JonR>well, I'm back within acceptable norms, so I'd rather spend my time working on my SoC project or goofing off on the net, than experimenting with ssh servers. So thanks, but no thanks
22:29<TheFirst>until apache starts getting loaded...that's what prompted me to start messing with things....even with cakers config it still kills a 'node after a while
22:44<JonR>something's fucked with my fastcgi setup.
22:45<JonR>I have 13 trac.fcgi processes. or threads. or something.
22:49<TheFirst>config problems...
22:51<JonR>got any good ideas for the FastCgiConfig values?
22:51<TheFirst>fix them once and they go away...or kill the procs and you've got a bandaid for a while...which is gonna take more time in the long run?
22:51<TheFirst>not mess with fcgi w/ apache
22:59|-|VS_ChanLog [] has left #linode [Rotating Logs]
22:59|-|VS_ChanLog [] has joined #linode
23:00|-|Battousai [] has joined #linode
23:24|-|Redgore [~Redgore@] has quit [Ping timeout: 480 seconds]
23:29|-|JonR [] has left #linode [Leaving]
23:30|-|Redgore [~Redgore@] has joined #linode
23:41|-|afv-13 [] has quit [Quit: leaving]
---Logclosed Tue May 30 00:00:02 2006