Back to Home / #linode / 2012 / 04 / Prev Day | Next Day
#linode IRC Logs for 2012-04-01

---Logopened Sun Apr 01 00:00:01 2012
---Daychanged Sun Apr 01 2012
00:04<KyleXY>retro|blah: hardly, lol
00:04<retro|blah>KyleXY: Yeah. I was trying for first on April 1st (ET), not first after VS_ChanLog cycled
00:05<KyleXY>retro|blah: *shrug*
00:05<retro|blah>I was hoping to hit it right on :00 too, but I missed it by 1s
00:05<retro|blah>Oh well
00:08-!-BarkariII [BarkerJr@2002:1880:41b:1:982d:6d4e:5a8a:cdd7] has joined #linode
00:09-!-BarkerJr [BarkerJr@2002:1880:41b:1:982d:6d4e:5a8a:cdd7] has quit [Read error: Connection reset by peer]
00:10-!-arooni-mobile [~arooni-mo@] has quit [Ping timeout: 480 seconds]
00:10-!-Bass10 [] has quit [Read error: Operation timed out]
00:15-!-Katana [] has quit [Quit: How to start a project destined to fail: carry an attitude of "I understand it, that's good enough"; Keep it simple, stupid.]
00:15-!-Katana [] has joined #linode
00:15<Katana>fffff. it's cached still.
00:15<Katana>(╯°□°)╯彡/(.□ . \)
00:19<linbot>(╯°□°)╯彡 ┻━┻
00:22<chesty>every time you wonder if the cache has expired, you've lost the game
00:22-!-flabbergaster [] has joined #linode
00:23-!-mcinerney [] has joined #linode
00:25-!-aogier [] has joined #linode
00:25-!-DreamPhysix [] has joined #linode
00:25<flabbergaster>Hi all, I'm having trouble inserting a binary string into mysql, using php/mysql or mysql (via command line). I'm able to insert binary data using php on my local machine, but when I run hte same script on the server, the data that I see inserted is a bunch of question marks. Any idea what I should look into?
00:25<DreamPhysix>Can someone help me with a security issue?
00:26<linbot>If you have a question, feel free to just ask it -- someone's always willing to help. If you don't get a response right away, be patient!
00:26<DreamPhysix>Someone was sending thousands of requests to one of my websites which spiked CPU usage and I want to parse the log file to just show me everything by that IP. What's the best way to do that?
00:26<DreamPhysix>grep IP_ADDRESS > output.txt ?
00:26<Kyh_>DreamPhysix: "grep <ip> log.log"
00:27<bdube>fgrep is handy there
00:27<Kyh_>flabbergaster: does php/mysql have a "input type" thing?
00:27<DreamPhysix>well i need to find the log on the server then
00:27<DreamPhysix>i have it locally
00:27<Kyh_>DreamPhysix: then do that
00:28<flabbergaster>Kyh_: not sure what you mean. the column type is varchar, and the collation is utf8
00:28<bdube>flabbergaster: you want binary bits in varchar? that doesn't sound right.
00:29<flabbergaster>bdube: i know, what's weird though, is that it works fine locally
00:29<aogier>My arch linux linode is missing a lot of basic utilities like 'hostname'. I don't really want to 'pacman -S base' because that will get me a new kernel etc. Anyone know what packages linode obsoletes?
00:30<bdube>flabbergaster: are you supplying the data the same way to the remote system as you do locally?
00:30<flabbergaster>bdube: should i be using the binary type? was trying to avoid rebuilding a big table (has an extra varchar field i can reuse), but may end up rebuilding it anyways
00:30<DreamPhysix>Does anyone mind briefly looking at my HTTP access log and seeing what the user was doing? It just looks like a bunch of scripted HTTP requests to me
00:30*DreamPhysix pastes
00:30<linbot>Maybe not a verb. But you can do it here:
00:31<flabbergaster>bdube: yeah, the data is saved as a string in the script
00:31<flabbergaster>and am using the standard mysql functions
00:31<rnowak>are you sure the data is actually wrong, and not just a display issue in your terminal?
00:31<aogier>flabbergaster: maybe the data is correct, but it is output differently
00:31<rnowak>screw you linbot
00:32<flabbergaster>dammit rnowak, you may be right
00:32<rnowak>flabbergaster: create a hash of it and compare to local
00:32<flabbergaster>okay, i will try that out now
00:33<DreamPhysix>My paste is too big for :(
00:33<retro|blah>aogier: Apparently net-tools (which would provide hostname) is deprecated on arch, and is superseded by yp-tools
00:34<aogier>retro|blah: yeah, i think the minimal image that linode provides just included net-tools
00:34<aogier>i'd still like to know though, is it documented somewhere what binaries linode provides?
00:35<flabbergaster>rnowak: you're right!
00:35<aogier>like, would the world blow up if i 'pacman -S linux'ed?
00:35<DreamPhysix>Should I report this IP address to the Internet provider if it sent 116k requests to my server in under 24 hours?
00:35<flabbergaster>the data is correct, my display is messed up
00:35<flabbergaster>argh, thx terminal
00:35<linbot>DreamPhysix: IP:; rDNS: None; City: Hartford; State: Wisconsin; Postal code: 53027; Country: United States;
00:35<rnowak>flabbergaster: ;)
00:36<retro|blah>aogier: It shouldn't blow anything up if pacman pulls in a kernel image. Your Linode would ignore it, unless you decided to set up pv-grub
00:36<linbot>joshie_: IP:; rDNS: None; City: Moscow; State: Moscow City; Country: Russian Federation;
00:36<retro|blah>Normally you boot from Linode's kernel, which is not managed by your distro's package manager
00:36<aogier>retro|blah: so basically linode just runs its own linux kernel, on its own filesystem?
00:37<aogier>there aren't traces of it on my disk image?
00:37<flabbergaster>...and of course changing the encoding on my terminal fixed it. thanks rnowak and bdube!
00:37<retro|blah>aogier: It would not be on your disk image, correct
00:37<retro|blah>Probably it's stored on the host your linode is running on
00:37<aogier>alright, thanks for the help
00:38-!-flabbergaster [] has quit [Quit: ajax IRC Client]
00:40-!-DreamPhysix [] has quit [Quit: DreamPhysix]
00:44-!-h4 [] has joined #linode
00:47-!-diimdeep [~diimdeep@] has joined #linode
00:47<h4>Anyone here? I'm looking for IPv6 speed tests.
00:48<h4>Speedtest page has only tests for IPv4
00:49<Kyh_>you're right, they dont have AAAA records
00:50<KyleXY>don't think the hosts themselves have IPv6 connectivity
00:50*Kyh_ pokes Perihelion or someone
00:50<KyleXY>Kyh_: ^^ he
00:51<KyleXY>Kyh_: I don't think they have any ipv6 addresses brought up for the hosts themselves,
00:51<@Perihelion>The nameservers do
00:51<@Perihelion>You can use ns1-5
00:51<KyleXY>Perihelion: referring to
00:51<retro|blah>That would be good for latency testing only though wouldn't it?
00:52<KyleXY>the xenhosts,
00:52<@Perihelion>Yeah, pretty much
00:52<Kyh_>Perihelion: cant use those for speed tests tho
00:52<KyleXY>Yay for being right, ♥
00:52<@Perihelion>Not speed tests, no. But testing latency yes.
00:53<auraka>h4: what location do you want to test?
00:54<KyleXY>auraka: I imagine they probably wanted to poke each datacenter to see which was fastest for the
00:54<h4>I'm in china and using native ipv6 connection
00:56<auraka>h4: you can trace to 2400:8900::f03c:91ff:fedf:ef05 for speed test anymore
00:56<h4>Ok, thank you
00:57<Kyh_>no in tokyo either
00:58<h4>It seems that the datacenter in CA is much faster... So strange
00:59<auraka>h4: routing in Asia is wonkey
01:00<h4>I'll have a try...
01:03<Kyh_><3 spooning
01:03<Kyh_>On, not that kind of spooning
01:05-!-alx- [~alx-@] has joined #linode
01:05-!-alx- [~alx-@] has quit []
01:05<h4>I traced it again. Then the latency was half of my first trace.
01:06<h4>Thanks a lot
01:06*joshie_ tests Perihelion's latency
01:09<auraka>joshie_: are you trying to say she isn't smart?
01:10<@Perihelion>I mean, have you spoken to me recently?
01:10<purrdeta>hmm I have.
01:11<rnowak>I tried to yesterday but you didn't hear me through the window
01:11<@Praefectus>last time i spoke to Perihelion she sounded like the teacher on charlie brown
01:12<praetorian>rnowak: you also said it was 3am in the morning and she was asleep
01:12<@Perihelion>I haven't seen Praefectus in a while. Pretty sure he's eternally cancering.
01:12<joshie_>i think it's weird you related latency with intelligence
01:12<@Praefectus>hey if i had cancer, the doctors would have seen it by now
01:12<joshie_>i was more making a sexual joke
01:12<@Perihelion>You could at least buy me dinner first.
01:12<rnowak>are you drunk again?
01:13<rnowak>it was pretty pathetic last time, maybe you should just go to bed
01:13<@Perihelion>I am not.
01:13<rnowak>not you, joshie_
01:13<@Perihelion>Oh, I was like "Ow, my pride :<"
01:13<joshie_>Perihelion where would you like to go?
01:13<@Perihelion>ruths chris
01:13<@heckman>it was delicious, js
01:13<@heckman>best steak I think I've ever had
01:14<@Perihelion>the one in ac isnt as good as the one in dc
01:14<@heckman>The one in Columbia was amazing
01:14<auraka>rnowak: drank some hatorade?
01:14<rnowak>auraka: nop
01:15<@Perihelion>the one in columbia scalps the crap out of you
01:15-!-h4 [] has quit [Remote host closed the connection]
01:17<praetorian> best ever
01:18<@Perihelion>I want sea monkeys again
01:18<praetorian>so oldschool.
01:21<@Praefectus>rdg in houston has the best steak i ever had
01:22<praetorian>looks too fancy to be a nice steak place
01:22<@Praefectus>the coffee steak is pure awesome
01:22<linbot>New news from forums: SFTP permissions advice in Web Servers and Web App Development <>
01:24<purrdeta>hmmmmm steak...
01:24<purrdeta>I think I'll go eat ice cream out of the container.
01:25<Kyh_>hrghrhg ftp
01:26<linbot>Please use SSH/SCP/SFTP/rsync-over-ssh instead of FTP:
01:31<joshie_>Perihelion i found u
01:33-!-debug74 [] has joined #linode
01:33<debug74>Hi to everyone
01:33<joshie_>oh google
01:34<joshie_>u so funnay
01:34<joshie_>rror 502 (Server Error)!!1
01:34<joshie_>Error 502 (Server Error)!!1*
01:34-!-seanh-ansca [] has joined #linode
01:34<debug74>I have a jail chroot problem , anyone can help?
01:35<joshie_>chroots are highly dangerous, i shall direct your question to XReaper
01:35-!-dhubbard [] has quit [Read error: Operation timed out]
01:36<linbot>If you have a question, feel free to just ask it -- someone's always willing to help. If you don't get a response right away, be patient!
01:36-!-metasansana [~metasansa@] has quit [Quit: :O]
01:37<debug74>I have set up a jailed chroot to a folder on my machine, /usr/local/chroot
01:37<debug74>I installed the rssh
01:37<debug74>configured it to /usr/local/chroot
01:38-!-nmudgal [~tracker@] has joined #linode
01:38<debug74>I have added a user(name is user1) with the home directory to /usr/local/chroot/user1
01:38<debug74>I created a .ssh folder and a authorized_keys file in user1's home folder
01:39<debug74>added my public keys to the file . I sftp using user1 and it said server refused my key
01:40<@heckman>have you checked your auth log?
01:40<debug74>my public keys is used on the same machine for other user which has their home directory set to /home/<user>
01:41<debug74>my machine is on Centos 5
01:41-!-rurufufuss [] has joined #linode
01:41<debug74>I have checked the secure log file
01:41<debug74>it says trying public key file /usr/local/chroot/user1/.ssh/authorized_keys
01:41<debug74>no permission errors
01:42<debug74>no errors. it just simply did not accept my keys
01:42<debug74>sshd_config logging have been set to DEBUG
01:42<debug74>no errors logged
01:45-!-ronkrt [] has joined #linode
01:45<ronkrt>i just deployed a linode with a stackscript
01:45<ronkrt>do i have to run the stackscript or is it automatic?
01:45<@heckman>runs on first boot
01:47<ronkrt>ahh awesome
01:47<ronkrt>i can watch it with tail -f shellstack.log
01:50-!-seanh-ansca [] has quit [Quit: Leaving.]
01:52<dopey>debug74: if its not something simple, like the user not owning ~/.ssh/ and ~/.ssh/authorized_keys, or either the dir or file being group writable, then one way to debug is to start sshd in debug mode ( /usr/sbin/sshd -D -d -p 8022 )
01:53-!-dhubbard [] has joined #linode
01:56<@heckman>Seems people have issues with rssh on CentOS 5.6
02:01<auraka>rnowak: ping
02:02<rnowak>auraka: pong
02:02-!-nmudgal [~tracker@] has quit [Ping timeout: 480 seconds]
02:03<auraka>rnowak: i believe you have a disdain for cpanel, did you happen to see that they are now forcing a third party SEO service into their panel....pretty interesting
02:04<rnowak>I don't really give a crap about it, I just lulz a bit at it when given a chance
02:04<Kyh_>auraka: lol
02:04<Kyh_>requires moar IPs I guess
02:04<auraka>rnowak: enjoy the lulz
02:05<rnowak>very nice
02:05-!-Guest181 [] has quit [Quit: This computer has gone to sleep]
02:12<ronkrt>is there an easyway to find out why a stackscript may or maynot have failed?
02:13-!-nmudgal [~tracker@] has joined #linode
02:18<rnowak>auraka: how's the fusionio treating you?
02:32<Kyh_>ronkrt: logs?
02:32<ronkrt>just finished rebuilding it, to try again
02:33<ronkrt>let you know after i get the next boot up done, all it said last on the logs was starting wp install
02:38<ronkrt>yes!! just got networking inot meh game w00t
02:50-!-jgornick [] has quit [Quit: jgornick]
02:55-!-fayimora [] has quit [Quit: Busy…..zzzzz]
02:59-!-fayimora [] has joined #linode
03:05<ronkrt>has anyone used ?
03:07-!-dr_jkl [] has quit [Quit: fuck irc]
03:08<ronkrt>now i cant get into my box
03:12<ronkrt>that stackscript keeps hanging on triggering wordpress install
03:13<Kyh_>what does LISH say?
03:14-!-Zr40 [] has quit [Quit: leaving]
03:14-!-Zr40 [] has joined #linode
03:29<ronkrt>Lish let me in, when i read the .log
03:29<ronkrt>it says it was initiating the wordpress install but its stuck there
03:29<ronkrt>i tailed it with tail -f shellstack.log
03:31<retro|blah> hm.
03:33<XReaper>Hello World
03:33<ronkrt>hi XReaper
03:33<XReaper>cleaning my car :(
03:33<XReaper>(i wasn't the one who made the mess...)
03:34-!-sivy_ [] has joined #linode
03:34<XReaper>too bad linodes aren't THAT awesome
03:40-!-sivy [] has quit [Ping timeout: 480 seconds]
03:40<ronkrt>does the linode made wordpress stack us apache?
03:41<XReaper>ronkrt: i would think so
03:41<Zr40>see line 10
03:41<XReaper>nginx is a pain to write rules for
03:41<Zr40>I would disagree
03:41<Zr40>I'd say it's about as hard as for Apache
03:42<XReaper>I found my rules off the net
03:42<XReaper>wordpress comes with apache rules by default
03:42<XReaper>.htaccess bull
03:42<Zr40>so without wordpress-provided apache rules, you would have to write your own
03:42<Zr40>so this isn't an apache benefit, but a wordpress convenience
03:44<ronkrt>so even if i get a nginx wordpress stack running, i'll have to edit the rules alot?
03:44<Zr40>not a lot, you would only have to do it once
03:44<rnowak>!to ronkrt alot
03:44<linbot>ronkrt: |
03:44<Zr40>I like this alot
03:44<rnowak>a rule alot
03:45<Zr40>do you like alot a lot too?
03:46<rnowak>a lot of alots are likable a lot
03:47<Zr40>rnowak: I've never tried to edit a rule alot. Does the rule alot like that a lot?
03:49<rnowak>a lot of rule alloted alots allot rules a lot to rule alots to rule alots a lot
03:50<EugeneKay>What does alot of alots look like?
03:50<rnowak>so you see, they just want to be in charge
03:50<Zr40>sad alot :(
03:52<Zr40>EugeneKay: I hope it doesn't look like this...
03:52<EugeneKay>It looks alot like I would expect
03:54<@mikegrb>mmm cake
03:54<Zr40>alot of cake!
03:54-!-hl [] has quit [Remote host closed the connection]
04:04-!-vikram [] has joined #linode
04:05<vikram>Can anyone help me here please
04:05<linbot>If you have a question, feel free to just ask it -- someone's always willing to help. If you don't get a response right away, be patient!
04:08<vikram>We have developed a web app using GWT+MyBatis+Spring+MySQL. Need to know which plan will suite us?
04:08<Kyh_>er, how much RAM are you using under peak load?
04:09<vikram>Thanks Kyh for response. We are a web startup from India, And going live for the first time
04:10<Kyh_>Well, I have no idea, you should stress test it locally first before going live.. but if you start with a 512, you can simply upgrade, which takes ~10-20 minutes or so of downtime (mayb eless/more?) when you get migrated to a bigger plan
04:10<Kyh_>There's also the load balancing service too
04:14<vikram>ok.Right now we are on Windows Platform. And I dont have much hands on in Linux. So which Linux version to choose, and how we will get the compatible Tomcat and MySQL ?
04:14<Zr40>also, make sure to test your systems to see if you don't run in any unexpected capacity problems
04:15<Zr40>some badly written or misconfigured software tend to use a lot of memory under high load
04:16<Kyh_>er, you install them.. as to which distro.. well, that's a hard question. I'd say ubuntu/debian
04:17<vikram>is there any step by step guide to do all the required configuration other than
04:18<Kyh_>er, if it's not in the library, you could check google :P
04:19<chesty>vikram: it's going to be 1000 times better for you if your development environment matches your production environment.
04:20<vikram>but I think Linode doesn't provide the Window platform for deployment, am I correct?
04:20<Zr40>correct, it's Linux only
04:21<Kyh_>... yes, linode is linux only
04:21<rnowak>so what os is failnode running on?
04:27-!-Bartzy [~bar@] has joined #linode
04:27<Bartzy>How do I check information about a public key in openssl ?
04:27<joshie_>ey you
04:27<joshie_>ey ey
04:28<joshie_>i do not remember that off by heart
04:28<Kyh_>Bartzy: er, man openssl?
04:28<Bartzy>Kyh_: I can't find anything about public key info.
04:28<joshie_>le google
04:28<Kyh_>I imagine it'll be something about the certificate information, then
04:28<Bartzy>Kyh_: It's not a certificate, it's a public key.
04:28<Kyh_>since there aren't "just" "public keys"
04:29<Bartzy>it's not signed or something.
04:29<Bartzy>Yes there are
04:29<Kyh_>sure you don't mean gnupgp/ppg?
04:29<Zr40>that's not openssl
04:29<Zr40>that's ssh
04:29<Kyh_>right, there's ssh, not openssl
04:30<Bartzy>Kyh_: Still a regular public/private key.
04:30<Bartzy>I need to check how many bits a public key is
04:30<Bartzy>Sorry got disconnected.
04:31<Zr40>try ssh-keygen -l
04:31<Kyh_>sorry, you were saying openssl, so I was assuming you were talking about SSL certs, heh
04:32<Zr40>Bartzy: so, ssh-keygen -l -f path to public key
04:33<Zr40>I've got a failing cron job, but running the command myself succeeds. How would I debug this?
04:34<rnowak>how does it fail?
04:34<Zr40>exit status 1, according to crond.log
04:34<praetorian>so i wonder if i will get an april fools day invoice from linode
04:34<rnowak>nothing else?
04:34<praetorian>and if it is a fool :o
04:34<Kyh_>note that crons don't get the same shell environment, etc
04:34<rnowak>likelyhood is that the environment isn't the same for both, and using relative paths etc may not work
04:34<Zr40>absolute path
04:38-!-vikram [] has quit [Remote host closed the connection]
04:39<Zr40>does cron store stderr output somewhere?
04:41<rnowak>you should redirect it somewhere, but iirc it would be sent in the email to root unless you've disabled that
04:48<Zr40>is the only way to get stderr output sending it by email?
04:52<Zr40>how odd. it's trying to load a gem it shouldn't. (it should only be loaded when running unit tests)
04:52<Zr40>I don't understand why it's trying to do that with cron, but not when I run it myself
04:53-!-gerryvdm [] has joined #linode
04:54<rnowak>check the code around where it errors out? maybe something in the environment triggers it
04:55<gerryvdm>anyone familiar with PHP-FPM? I suppose there is no way to spawn processes with different user credentials without adding an extra pool?
04:57<Zr40>...great. Now it's suddenly reproducible
04:59<Zr40>okay, it was due to a wrong cd path
04:59<Zr40>note to self: *always* use set -e
05:19<ronkrt>do i have to manually start apache after the linode wordpress stack setsup?
05:27-!-stafamus [] has joined #linode
05:38<diimdeep>hello, why free -m show that only 495 total memory available but in linode manager set to 512 ?
05:39<Zr40>495 is the total memory available to applications
05:39<Zr40>the remainder is used by the kernel to work
05:39<Zr40>it is not included with the 495 MB because it must always be in memory
05:40<diimdeep>thanks, your answers really helpfull
05:47-!-SelfishMan [] has quit [Quit: meh]
05:51-!-hipsterslapfight [] has joined #linode
05:54-!-SelfishMan [] has joined #linode
06:09<Zr40>what does the 'migration/0' kernel process do?
06:10<joshie_>sounds VPS related
06:11<Zr40>because top shows 15174 minutes of CPU time for that process (not currently increasing). Uptime is 12 days 17 minutes, which is 17297 minutes
06:11<Zr40>I'm not having a problem (I think), I'm just curious
06:11<marius>So I'm playing around trying some new stuff out, so I decided to do a lode balancer, with two "front end" nodes and a backend one, but I must be doing something wrong websites bottleneck appears to be with MySQL (which is to be expected if it had any load, but the real stop is a 28second wait time for it to connect to the sql database in the first place)
06:12<marius>So how would I go about troubleshooting what could be causing this insane connection time?
06:12<Zr40>try using the command line mysql tools from one of the front end nodes to connect to the mysql server
06:15<marius>yeah, that too takes for ever
06:15<Zr40>how are you connecting?
06:15<marius>What do you mean ?
06:15<Zr40>by hostname or by ip?
06:16<Zr40>this is a wild guess, but try this on the server running mysql
06:16<Zr40>'host <front end ip>'
06:16<marius>I'm thinking it might be related to my firewall rules ?
06:17<marius>(I've got INPUT BLOCK with a single accept line for port 3306 from my front end)
06:17-!-robzilla [] has joined #linode
06:17<Zr40>perhaps. does host also take around 28 seconds before failing or succeeding?
06:18<marius>well it's all going by local IPs, but running host <ip of front> on the backend times out
06:18<Zr40>it could be that mysql is trying to resolve the hostname of the connecting client
06:19<marius>it dones't take 28 seconds ot tell me it fails though
06:23<marius>well I'll be damned
06:23<marius>that solved it right up straight away
06:24<marius>peculiar how it failed to host lookup even when I added the ip to the hosts file as well o_O
06:25<Zr40>some DNS libraries only use the DNS server, not /etc/hosts or other means of resolving names
06:26<marius>well that's silly of them :P
06:27<marius>oh well, not like I need the lookup any way, only pre-approved hosts can connect any way so I can recognize them easily enough
06:27<ronkrt>i cant get a single stackscript to work ;(
06:27-!-saikat [] has joined #linode
06:27<marius>Getting any errors ?
06:28<ronkrt>the one i wanted to use is getting stuck after processing wordpress install
06:28<ronkrt>other ones havnt had a log
06:31<ronkrt>wow go figure
06:31<ronkrt>i got 1 to finally go threw
06:34-!-Ricki [~Ricki@] has joined #linode
06:39<ronkrt>how do i get my user public ssh key?
06:42-!-Zr40 [] has quit [Quit: leaving]
06:42-!-debug74 [] has quit [Remote host closed the connection]
06:55-!-hipsterslapfight [] has quit [Read error: No route to host]
06:55-!-hipsterslapfight [] has joined #linode
07:09*azaghal thinks today is a very dangerous time of year to ask about anything\
07:09<ronkrt>these stackscripts are really pissin me off ;(
07:16-!-Bass10 [] has joined #linode
07:25-!-Vikram [] has joined #linode
07:25-!-derek [~derek@] has joined #linode
07:27<Vikram>Hi All, I just installed the MySQL on Ubuntu 10.04. Installation is successfull. Now when I am trying to access the mysql client, I am getting error "ERROR 1045 (28000): Access denied for user 'root'@'localhost' (using password: YES)"
07:29-!-atula [] has joined #linode
07:30<rnowak>solution: provide the correct password
07:32<Vikram>I feel its realted to some privileges . When I am typing mysql also, its gives same error "ERROR 1045 (28000): Access denied for user 'root'@'localhost' (using password: YES)"
07:32-!-Zr40 [] has joined #linode
07:32<rnowak>how are you providing the password?
07:32<rnowak>on the command line or when it prompts?
07:32-!-derek [~derek@] has quit [Quit: Leaving]
07:33-!-derekyang [~derekyang@] has joined #linode
07:34<Vikram>when it prompts
07:35<rnowak>are you providing the password for your system's root account, or mysql's root? if the latter, you'll need to reset the password as it seems you've forgotten it :p
07:38-!-DephNet[Paul] [] has joined #linode
07:39-!-maku`off is now known as mkasu
07:39<ronkrt>how can i make a stackscript log if it doesnt?
07:41<Vikram>Thankls ronkrt for Input
07:42<ronkrt>you mean rnowak ?
07:42<rnowak>log what ronkrt?
07:43<Vikram>Yes sorry its rnowak
07:43<ronkrt>rnowak, i dont know, when i tried to stall the WP Optimized stack it had a shellstack.log that i could tail
07:44<ronkrt>however it kept getting stuck and i want to try another, but it doesnt create a .log i can tail
07:45<rnowak>the logging information was provided by the shellstack recipes
07:46<ronkrt>where can i find it on a debian install using the my better wordpress and wp supercache?
07:46<rnowak>find what?
07:46<ronkrt>the shellstack log
07:47<rnowak>you implied above that you're no longer using the stackscript that uses shellstack
07:47<ronkrt>i might as well give the f up on these stacks
07:48<rnowak>they are shell scripts with some fluff; UDFs
07:50<ronkrt>i would like to get this, working but its epic failing, along with the wp optimzed one which is another i'd like to load, not even the linode wp one works ;(
07:53-!-TimTim [] has joined #linode
07:54<rnowak>how does the linode one not work? what happens?
07:54<ronkrt>lemme launch it again
08:00-!-TimTim [] has quit []
08:01<ronkrt>ok, it might be working
08:01-!-TimTim [] has joined #linode
08:01-!-fayimora_ [] has joined #linode
08:02<ronkrt>this time it worked
08:03<ronkrt>how can i copy the contents of and all sub directorys of a directory up 1?
08:05-!-TimTim [] has quit []
08:05-!-fayimora [] has quit [Ping timeout: 480 seconds]
08:05-!-fayimora_ is now known as fayimora
08:05-!-TimTim [] has joined #linode
08:07<ronkrt>actually got it working awesome
08:08-!-azaghal [~azaghal@] has quit [Remote host closed the connection]
08:08-!-azaghal [~azaghal@] has joined #linode
08:08-!-TimTim [] has quit []
08:09-!-TimTim [] has joined #linode
08:15-!-JM [] has joined #linode
08:15-!-TimTim [] has quit []
08:18-!-Bass10 [] has quit [Read error: Operation timed out]
08:27-!-wkl [~wkl@] has quit [Quit: wkl]
08:29-!-Vikram [] has quit [Remote host closed the connection]
08:38<staticsafe>heh Freenode's April Fool's joke was hilarious
08:38<dwfreed>staticsafe: have you figured out the puzzle?
08:39<GLaDOSDan>What was freenode's?
08:40<GLaDOSDan>Reddit's is one of the best so far imo
08:40-!-gerryvdm [] has quit [Quit: Linkinus -]
08:41-!-hipsterslovelegos [] has quit [Quit: ajax IRC Client]
08:43<GLaDOSDan>I love how the top is in comic sans
08:45<staticsafe> - lol at their april fools
09:12-!-goose [] has joined #linode
09:16-!-hipsters_ [] has joined #linode
09:17-!-hipsterslapfight [] has quit [Read error: No route to host]
09:17-!-hipsters_ is now known as hipsterslapfight
09:21-!-dubenstein [~dubenstei@] has quit [Remote host closed the connection]
09:21<linbot>New news from forums: Blocked outbound packets in Linux Networking <>
09:22<Katana>xkcd's is interesting. hasn't unraveled yet. something to do with a waldo.js i think
09:24<Katana>check the page source
09:43-!-goose [] has quit [Quit: Staying alive can kill you, it's taken years off of my life.]
10:04-!-wkl [~wkl@] has joined #linode
10:13<ronkrt>how can i make sure mode rewrite is running?
10:15<ronkrt>nvm got that
10:24-!-goose [] has joined #linode
10:38-!-TimTim [] has joined #linode
10:42-!-teacup [] has joined #linode
10:51-!-derekyang [~derekyang@] has quit [Remote host closed the connection]
10:51-!-gerryvdm [] has joined #linode
10:51<gerryvdm>is it better to run php-fpm over a unix socket than tcp or are there other drawbacks in that case?
10:52<XReaper>sockets > tcp
10:53<numk>win 1
10:53<dwfreed>numk: lose 1 :)
10:53<numk>i was changing irssi window qq
10:53<dwfreed>numk: I know, I use irssi too
10:54<gerryvdm>thx XReaper, and is there a recommend location to put these sockets?
10:54<gerryvdm> /tmp?
10:54<praetorian>i found the 2 irssi users of the world
10:54<XReaper>doing a /ctcp version on me is... lolz
10:54<XReaper>gerryvdm: the defaults should be fine
10:54<numk>what's wrong with irssi
10:54<gerryvdm>the defaults are tcp :)
10:54<XReaper>it does have defaults
10:54<numk>11:07 -!- Irssi: Uptime: 415d 15h 29m 41s
10:54<XReaper>When i installed php-fpm it had sockets by default
10:54<XReaper>that and nginx
10:55<praetorian>numk: nothing
10:55<XReaper>but i do run
10:55<linbot>The Romans used Arch. Worked out GREAT for their civilization.
10:55<gerryvdm>mine has tcp by default, via apt on ubunu 11.10
10:55<XReaper>i don't see anything wrong either way
10:55<gerryvdm>whats your default location then out of interest?
10:55<XReaper>*opens ssh window*
10:55-!-seanh-ansca [] has joined #linode
10:56<gerryvdm>sorry for bothering :)
10:56<dwfreed>gerryvdm: somewhere in /var/run should be fine
10:56<XReaper>server unix:/var/run/php-fpm/php-fpm.sock;
10:56<gerryvdm>will do, thx both
10:56<XReaper>/var/run is where that stuff belongs
10:56<dwfreed>gerryvdm: just make sure nginx or apache can write to it
10:57<XReaper>i'd assume sockets are more secure
10:57<gerryvdm>not sure why tcp is the default then
10:58<XReaper>ubuntu is weird
10:58<Zr40>sockets only work locally while tcp can be used to have a front end proxy connecting to the back end app servers
10:59<XReaper>but locally its fine
10:59<Zr40>I would also assume sockets have less overhead than local tcp
11:00<Zr40>though you wouldn't notice that unless the app responds instantly
11:01<A-KO>you mean named pipes?
11:01<Zr40>unix sockets
11:01<rnowak>well, loopback is quite efficient
11:01<rnowak>so the "less overhead" is probably not so much less
11:02<dwfreed>A-KO: there's unix sockets and named pipes; sockets are bi-directional, named pipes aren't
11:02-!-squircle [~squircle@2607:f2c0:a000:138:129a:ddff:fe50:5421] has joined #linode
11:03<Katana>dwfreed: is there a perf difference, out of curiosity?
11:03<Zr40>Katana: there's a behavior difference
11:03<dwfreed>Katana: not really
11:03<Katana>dwfreed: so, too-minor-to-give-a-damn
11:03<dwfreed>Katana: yep
11:04<dwfreed>A-KO: also, named pipes can only have 1 writer and 1 reader; unix sockets are just like TCP sockets; they have one listener, and as many connections as you'd like
11:04<gerryvdm>fpm sockets working, finally completely removed apache out of the stack
11:04<Zr40>dwfreed: actually, multiple listeners
11:05-!-squircle [~squircle@2607:f2c0:a000:138:129a:ddff:fe50:5421] has quit [Remote host closed the connection]
11:05<Zr40>can't do that with TCP sockets :)
11:05<XReaper>gerryvdm: nginx?
11:05<gerryvdm>yeah nginx
11:05<XReaper><3 nginx
11:05<gerryvdm>had it first reverse proxying php requests to apache
11:05-!-squircle [~squircle@2607:f2c0:a000:138:129a:ddff:fe50:5421] has joined #linode
11:05<XReaper>squircle: o/
11:05-!-seanh-ansca [] has quit [Quit: Leaving.]
11:06<Zr40>my setup: nginx --> unix socket --> 2 unicorns
11:06<squircle>\o XReaper
11:06<XReaper>O.o Zr40
11:06<A-KO>dwfreed: I guess I'm too used to the Windows way of doing things, where named pipes are more like sockets :P
11:06<Zr40>XReaper: ...what? :)
11:07<gerryvdm>any more straightforward things i can look into to improve performance of my nginx+fpm stack? already enabled apc
11:07-!-teacup [] has quit [Remote host closed the connection]
11:08<XReaper>nginx+php-fpm is pretty awesome
11:08<XReaper>caching is an option
11:09<Zr40>though blindly caching everything usually isn't a good idea
11:09-!-ronkrt [] has quit [Read error: Connection reset by peer]
11:09<gerryvdm>yeah gonna look into using memcached for stuff, but that requires changes in the code too
11:09-!-ronkrt [] has joined #linode
11:11<Gshock>gerryvdm: custom site? or some sort of cms
11:11<gerryvdm>well gonna put several project on the server, most written in a framework (symfony)
11:12<Gshock>Be careful with your APC settings imo. Make sure you toss apc.php somewhere you can see and tweak it.
11:13-!-teacup [] has joined #linode
11:17<ronkrt>is it possible to setup nginx to server on *.*.* and *.* threw /public_html ?
11:18<ronkrt>every domain / subdomain pointed over will work serving the same content so to speak
11:18<XReaper>Something like Server_Name _;
11:18<XReaper>i think...
11:18<XReaper>_ means any unmatched domain
11:20<gerryvdm>ronkrt: yes, you can add "default_server" to the listen directive
11:20<gerryvdm>listen 80 default_server;
11:20<ronkrt>if i ever get nginx to work then allow php5 & mysql i'll be golden :P
11:20<gerryvdm>but i personally redirect to one canonical domain name
11:20<gerryvdm>so google doesnt penalize me :)
11:21<teacup>there's tutorials in the library for setting up nginx, ronkrt
11:22<gerryvdm>this is what i do
11:24<Zr40>This pastebin [..] runs on a Linode 300.
11:26<XReaper>isn't that a bit outdated?
11:26<Whoop>Yes, the 512 used to be a 300 :)
11:26<teacup>quite extinct indeed
11:26<Zr40>with a 360 in between
11:26<Whoop>yeah and a 256 before iirc
11:26<XReaper>real old
11:26<Zr40>the 300 became 360 in december 2007
11:26<XReaper>i'm on a 512, its awesome
11:27<Zr40>same :)
11:27<XReaper>In Australia, $20/mnth gets you a 128MB with some providers
11:28-!-saikat [] has quit [Quit: saikat]
11:29<gerryvdm>in belgium, $20/mnth give you a shared hosting with 2,5GB HDD
11:30<XReaper>gerryvdm: awesome
11:30<XReaper>should totes take it
11:30<gerryvdm>with the leading belgian hoster, mind you
11:30<XReaper>Hence the appeal for linode
11:30<gerryvdm>actually, its worse, cause its €20
11:30<XReaper>honestly... other Xen providers cut corners
11:31<XReaper>others use OpenVZ
11:31<teacup>otherwise known as over-sell mania
11:32<Zr40>there's a Dutch provider offering a € 20 VPS beating the Linode 512 specs on everything except CPU
11:32<Zr40>I wonder what's the catch...
11:32<gerryvdm>datacenter in canada? :)
11:33<Zr40>nope, at the AMS-IX
11:33<Katana>*datacenter located in middle earth
11:33<gerryvdm>*in mordor
11:33-!-SnakeO [] has quit [Ping timeout: 480 seconds]
11:33<XReaper>Zr40: CPU means EVERYTHIGN
11:33<Zr40>ah, found a KVM logo.
11:33<XReaper>Some providers use SAN
11:34<XReaper>not local storage
11:34<XReaper>SAN == sloooow
11:34<dwfreed>Zr40: 20 Euro > 20 USD :)
11:34<Zr40>they mention 'storage servers'
11:34<numk>zzz SAN
11:34<XReaper>all you could use it for
11:36<squircle>$1 CAD > $1 USD *trollface*
11:36<rnowak>Zr40: which provider? tilaa?
11:36-!-teacup [] has quit [Remote host closed the connection]
11:36<XReaper>Tested an NY based host once...
11:36<XReaper>cpu alloc was shithouse
11:37<Zr40>rnowak: it's this:
11:37<rnowak>yeah, looking at it
11:38-!-eagles0513875 [] has quit [Read error: Connection reset by peer]
11:39<gerryvdm>is there any way to browse a webserver over ipv6 if your provider doesnt offer ipv6 yet for testing?
11:39<Zr40>yes, using a tunnel
11:39<XReaper>gerryvdm: tunnel server
11:39<XReaper> has em
11:40<rnowak>or w3m/links2/lynx on your linode ;p
11:40<XReaper>the ipv4 routing on ALLLLLLLLLL the enbpoints...
11:40<XReaper>goes via fremont
11:40<Zr40>how'd you mean?
11:40<gerryvdm>rnowak: yes but using my linode is cheating cause it would be loopback :)
11:41<rnowak>gerryvdm: got two linodes? :p
11:42<XReaper>Zr40: tokyo endpoint from a tokyo linode
11:43<Zr40>where's fremont?
11:43<XReaper>in the US
11:43<Zr40> that traceroute
11:43<XReaper>linode has servers there
11:43*rnowak facedesk
11:44<XReaper> <-- wherever that is
11:44<Zr40>might be san jose
11:44<XReaper>yeah... something like that
11:44<XReaper>shit routes
11:44<gerryvdm>how do you know that lynx used the ipv6 address over the ipv4?
11:44<XReaper>could it be intentional?
11:44<avenj>I assure you my tunnelbroke route to my v6 linode does not go through fremont.
11:45<XReaper>avenj: thats just the ipv4 tokyo endpoint
11:45<XReaper>via ipv6, 2ms away
11:45<XReaper>may as well just assign a /48 to my node and sub-allocate :/
11:46<Zr40>XReaper: if it helps, it also goes through sjc from here
11:46<XReaper>It gets to japan
11:46<XReaper>and bounces to san jose
11:46<XReaper>and back
11:49<XReaper>the footer on the paste site is oooooold
11:49-!-gerryvdm [] has quit [Quit: Linkinus -]
11:52<XReaper>traceroute6... 2ms hits same peering servers too
11:52<Zr40>which ip?
11:53<XReaper>i guess i'd use that to make a tunnel
11:55<XReaper>thats my endpoint
11:58<Zr40>that looks like ZF2
12:00-!-arooni-mobile [~arooni-mo@] has joined #linode
12:06<XReaper>iiNet is apparently allowing public access to their 6rd servers on ipv6 launch day shit
12:07<XReaper>(my isp)
12:07<XReaper>they bought out an isp who was offering ipv6 already...
12:08-!-Bryen [~bryen@] has joined #linode
12:09-!-lex [] has joined #linode
12:09-!-lex is now known as Guest261
12:12-!-shels [] has joined #linode
12:13-!-Guest261 [] has quit []
12:13-!-[1]robzilla [] has joined #linode
12:15-!-rurufufuss [] has quit [Remote host closed the connection]
12:15<Bartzy>How do I setup a init.d script to run after all other init.d scripts ?
12:17-!-shels [] has quit [Remote host closed the connection]
12:19-!-robzilla [] has quit [Ping timeout: 480 seconds]
12:19-!-[1]robzilla is now known as robzilla
12:23<@heckman>Bartzy: you could just not have it happen automatically and append it to your rcS file
12:25<Bartzy>heckman: ?
12:25<@heckman>Well, a line to run the script.
12:25<Bartzy>I want update-rc.d to handle it
12:25<@heckman>Well, I don't know of a way to do it last. Any time I've looked up doing it last I've seen instructions to add it to rcS
12:26<squircle>Bartzy: /etc/rc3.d/S99urmom?
12:26<squircle>there's probably some other S99's already, but...
12:27-!-alx- [~alx-@] has joined #linode
12:28<XReaper>IF the os even works that way
12:28<XReaper>In arch it isn't like that
12:30-!-alx- [~alx-@] has quit []
12:30<linbot>Linode's abuse contact is , as shown in the abuse contact info for the IP address in question. shows how to look this up yourself.
12:30<Bartzy>squircle: How do I do it with update-rc.d ?
12:30<Katana>Denyhosts email, whee. Looks like same datacenter
12:30<Bartzy>There's something about update-rc.d blah-script defaults 99
12:30<Bartzy>but it doesn't work.
12:30<linbot>Katana: IP:; rDNS: None; City: Mount Laurel; State: New Jersey; Postal code: 08054; Country: United States;
12:31<squircle>Bartzy: you don't even need update-rc.d, really... just make the file
12:31<squircle>otherwise, I'd refer you to `man update-rc.d` (because I dunno)
12:34<Bartzy>squircle: OK, thanks. I would prefer using update-rc.d
12:35*Katana slings abuse@ email heckman's way
12:35<squircle>Bartzy: well I just tried an example script with update-rc.d test enable, so...
12:36<Bartzy>ahh maybe enable is what I'm looking for and not 'defaults'
12:36<squircle>well, wait.
12:36<squircle>what are you trying to do?
12:37-!-arooni-mobile [~arooni-mo@] has quit [Ping timeout: 480 seconds]
12:38<squircle>Bartzy: if you're trying to get something to start last and shut down first, update-rc.d coolprogram defaults 99 01
12:44-!-ronkrt [] has quit [Ping timeout: 480 seconds]
12:44-!-Bartzy [~bar@] has quit [Ping timeout: 480 seconds]
12:50-!-vraa [] has joined #linode
12:56-!-Bartzy [~bar@] has joined #linode
12:57<@heckman>Katana: at home. :p
12:58-!-Dedalo [~Dedalo@] has joined #linode
13:02-!-mdcollins_ [] has joined #linode
13:03-!-robzilla [] has quit [Read error: Connection reset by peer]
13:03<XReaper>lol umad...
13:03<XReaper>nice hostname there
13:03<Katana>heckman: does this mean you'll give me food
13:03<Katana>i hear people have food at their homes
13:03*XReaper gives Katana some food
13:04-!-robzilla [] has joined #linode
13:04*XReaper gives Katana a katana
13:05<Katana>heckman: :C
13:06-!-mdcollins_ [] has quit [Read error: Connection reset by peer]
13:06-!-mdcollins [] has quit [Ping timeout: 480 seconds]
13:06-!-mdcollins_ [] has joined #linode
13:09-!-Perihelion [] has left #linode []
13:12*staticsafe gives Katana cookies
13:14-!-fatdruid [~wircer@] has joined #linode
13:15<@heckman>Katana: heh, I just got a fail2ban email about that one
13:16<mdcollins_>Have I been dc'ing all morning?
13:16<Katana>heckman: you too eh?
13:16<Katana>looks like a poisoned node.
13:17<mdcollins_>Comcast seems to be dropping packets >_<
13:19*heckman handles then goes back to getting his rear handed to him in BF3
13:19<dwfreed>heckman: heh
13:20<mdcollins_>Oy, I have the same problem in that game..
13:21<@heckman>My team is terrible which is the problem.
13:21<@heckman>Also, I'm currently working on all unlocks for recon my rate of fire is very limited. :)
13:26-!-fatdruid [~wircer@] has quit [Ping timeout: 480 seconds]
13:26<LiquidAtom>It's times like this I wish my system had SSDs instead of HDDs, copying over 400GB has taken in excess of 5 hours, and it's still got ~2 hours to go.
13:27<@heckman>shotgun to the face
13:27<Katana>heckman: Hope you've seen rendezook.
13:28<Katana>heckman: now try sniping someone instead of using an RPG. XD
13:29<@heckman>You've seen that video as well?
13:29<@heckman>Same situation, the guy snipes the dude out of the other jet and steals his jet
13:29-!-mdcollins_ [] has quit [Ping timeout: 480 seconds]
13:31<Katana>That's a bit harder to do in BF2
13:33<@heckman>I'm hoping to get my W/L ratio above 2.
13:34<dwfreed>heckman: good luck with that
13:34-!-diimdeep [~diimdeep@] has quit [Remote host closed the connection]
13:34<@heckman>Sitting at about 1.8 right now
13:37<gylt>this is the last straw,
13:37<gylt>I'm swearing off of freenode for good!
13:37<Zr40>there's often a straw shortage in early April
13:37<dwfreed>gylt: what day is it today?
13:38-!-Xenc [] has joined #linode
13:39<gylt>dwfreed: sunday.. why?
13:39<squircle>gylt: what day and what month?
13:39<Katana>gylt: full date, not day of week
13:39<gylt>guys, do $ date on your term
13:39-!-mramsden [] has joined #linode
13:40<squircle>gylt: it's april fools, silly
13:40<dwfreed>gylt: Sun Apr 1 17:40:14 UTC 2012
13:40<dwfreed>erm, meant to paste a link
13:41-!-zeade [] has joined #linode
13:41<EugeneKay>$ date +%s || 1333302068
13:41<EugeneKay>Looks like any other date to me.
13:42-!-mramsden [] has quit []
13:42<EugeneKay>Also, look like we'll hit 1333333333 today.
13:43<Zr40>these humans have an odd fascination with decimal sequences.
13:44<dwfreed>EugeneKay: well, by UTC time, it'll actually be tomorrow
13:44<EugeneKay>s/today/this sleep cycle/
13:45<linbot>EugeneKay: Yo mommas so stupid, she takes herbal penis enlargement pills. (790:4/11) [rummo]
13:47<dwfreed>2012-04-02 02:22:13 UTC-0000 == 1333333333
13:48<SleePy>What if shes not stupid. What if that actually works :|
13:49<Katana>nobody said she was taking it FOR that
13:52<SleePy>Just saiyan, this world is not black and white. There are too many shades of grey
13:52<squircle>linbot: urmom vote down 790
13:52<linbot>squircle: Voted down 790 [murom]
13:52<squircle>i agree with SleePy
13:55<squircle>also, random questIon: does anybody here live in BC and want 8 free BC Transit tickets? (I'd use 'em, but I live in Toronto, so...)
13:57<GLaDOSDan>Apr 1 19:49:40 sshd[29565]: Invalid user shit from
13:57<GLaDOSDan>sounds legit
13:57-!-seanh-ansca [] has joined #linode
13:57<CornishPasty>GLaDOSDan: nice :P
13:58<gylt>BC = ?
13:58<CornishPasty>squircle: I would but I live in the UK so...
13:58<squircle>gylt: british columbia (province of Canada)
13:58<gylt>oh, right right
13:58<squircle>*shrug* I figured I'd ask; it'll be a year before I go back and the fares will probably rise before then
13:58<CornishPasty>squircle: what were you doing all the way in BC?
13:58<squircle>CornishPasty: skiing
13:58<CornishPasty>Ooh fun
13:59<CornishPasty>Which mountain, squircle?
13:59<SleePy>GLaDOSDan is lying! He did that himself to appear cool!
13:59<squircle>whistler, of course
13:59<CornishPasty>Ah yes, I'd quite like to visit Banff
13:59<CornishPasty>Although I'm a bit fat now :(
13:59<squircle>I love banff; I grew up in Calgary, so I visited quite often
13:59<CornishPasty>Ah nice
14:00<squircle>s/grew up/lived for a few years/
14:02<CornishPasty>I grew up in the UK, but now want to leave
14:04<CornishPasty>squircle: smuggle me into Canada? :P
14:05-!-gerryvdm [] has joined #linode
14:13<dwfreed>grr, diagnosing network issues on my desktop that happen intermittently
14:16-!-jhanjon [] has joined #linode
14:16-!-blindwaves [] has quit [Quit: Leaving]
14:16-!-melfy [] has joined #linode
14:16<jhanjon>anyone using nagios on centos 6?
14:17-!-wkl [~wkl@] has quit [Quit: wkl]
14:17<squircle>CornishPasty: I'll trade you; always wanted to live in the UK
14:18<SleePy>It is so annoying to see Ubuntu/Debian having such old releases of Mysql. 5.1.58... I really want 5.6 but do not feel like building it from source...
14:19<CornishPasty>squircle: wooh!
14:19<CornishPasty>Btw squircle, our government is implementing big brother so they can spy on your internet usage and calls
14:25<LiquidAtom>CornishPasty: Yeah, it breaks the Human Rights Act and the Regulation of Investigatory Powers Act.
14:25-!-zivester [] has quit [Read error: Connection reset by peer]
14:25<LiquidAtom>But, because it'll be law, I guess that makes it okay! ^_^
14:25-!-zivester [] has joined #linode
14:26<squircle>CornishPasty: :| well, I could live with that. tunnel ALL the things.
14:26<squircle>I still love canada though
14:28-!-Silicis [] has joined #linode
14:28-!-Silicis [] has left #linode []
14:29<LiquidAtom>It makes me laugh that the Queen is going to announce it formally in her Jubilee speech, and it'll hopefully be enforced before the Olympics start.
14:33<staticsafe>hrm namebench tells me Google DNS is faster on average than the Linode nameservers :o
14:33-!-zivester [] has quit [Read error: Connection reset by peer]
14:33-!-zivester [] has joined #linode
14:35-!-nofxx [] has joined #linode
14:36<hawk>staticsafe: Yeah, because google knows what you will look up even before you do :p
14:38<nofxx>It's sunday, but for staff: Supposing a nosql multidatabase scenario, getting more common everyday, one could have say, 4 machines running DB and not using any outbound traffic. Could a plan where one exchange bandwidth for disk space happens in the future? heh
14:38<EugeneKay>IIRC, the Linode NSes attempt DNSSEC. Google's do not.
14:38<EugeneKay>nofxx - Proooobably not.
14:38<EugeneKay>Hosts only have so much disk space. There are extras available, but it's cheaper to just upgrade your plan.
14:39<EugeneKay>If you have large needs you should investigate S3 or a dedicated machine.
14:39<nofxx>eugene: yeah, the disk extras are really expensive, upgrade is cheaper. (can't autocomplete your nick sorry) heh
14:40<EugeneKay>S'okay, I have 'eugene' and 'EK' and some other things on highlight
14:40<bdube>it would have been only three more charact
14:40<nofxx>hmm... you need to go with elastic for that.. not possible to store db files on S3 and run the daemons on linode, is it?
14:41<EugeneKay>Possible, yes. Sane, no.
14:41<EugeneKay>If it's "nosql" you can investigate their hosted DB product
14:41<EugeneKay>I forget what it's called
14:42<dwfreed>EugeneKay: yes, Linode's resolvers are DNSSEC-validating, which makes them a bit slower (dig from a Linode, at the Linode resolvers, and you'll see "ad" in flags, which means the DNSSEC checks passed)
14:42<EugeneKay>That's what I was suggesting :v
14:42<bdube>DynamoDB or some such
14:43<EugeneKay>Also consider that Google DNS has a billion users hitting against it - they can cache doubleplusgood. Linode doesn't have quite so many.
14:43<squircle>dwfreed: i don't get ad in flags :(
14:43<dwfreed>squircle: wut?
14:43<staticsafe>i don't either
14:43<squircle>dwfreed: just tried what you said against; nothing
14:44-!-jhanjon [] has left #linode []
14:45<dwfreed>hmm, whacky; I see it against resolver1.dallas
14:45-!-zivester [] has quit [Read error: Connection reset by peer]
14:45-!-zivester [] has joined #linode
14:46<retro|blah>;; flags: qr rd ra ad; QUERY: 1, ANSWER: 1, AUTHORITY: 6, ADDITIONAL: 1 when querying against resolver2.newark
14:46<squircle>but... but...
14:46<dwfreed>retro|blah: did you ask for dnssec checking, though?
14:46<squircle>;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 6, ADDITIONAL: 0
14:47<retro|blah>Ah. No
14:47<dwfreed>wait, I bet I know what it is
14:47<dwfreed>retro|blah: you're running gentoo, right?
14:47<staticsafe>dig incompatibilites?
14:48<dwfreed>staticsafe: dig changes by ISC in newer versions
14:48<dwfreed>retro|blah: what version of bind-tools do you have?
14:48<squircle>ah, 9.7.3 here
14:48<staticsafe>DiG 9.7.3
14:49<dwfreed>I believe 9.8 enabled asking for dnssec by default
14:51-!-seanh-ansca [] has quit [Quit: Leaving.]
14:53-!-dubenstein [~dubenstei@] has joined #linode
15:04<Zr40>for 9.7.3, you can add +dnssec
15:04<synapt>A invoice notification on april fools day, too bad the bill isn't an april fools
15:04<Zr40>I believe invoices are always sent on the first
15:05<Katana>odd, i still don't have an invoice yet
15:05<Zr40>got mine 18 minutes ago
15:05-!-Ricki [~Ricki@] has quit [Ping timeout: 480 seconds]
15:05<synapt>Zr40: I know, but it's april fools day, so, you know
15:05<Katana>i don't have a balance either o.O
15:05<synapt>They probably mail out gradually over the day
15:05<Zr40>Katana: you get an invoice regardless of balance
15:06<Zr40>Katana: I got an invoice for deleting a linode :)
15:06<synapt>bulk-mailing == pretty fast way to get into automated blacklists/filters
15:06<Katana>Zr40: mm, yeah, but I would think right now i'd have a due balance
15:06<Zr40>Katana: hmm? referrals?
15:07<synapt>Katana: if you login to the site you'll probably see it
15:07<Katana>Zr40: none
15:07<Katana>synapt: manager says 0.00
15:07<synapt>ah, maybe they have it set to update the online system at the same time as the email, dunno
15:07*Katana shrug
15:07<Katana>I'll wait until tomorrow anyways, I guess
15:07<synapt>less overhead load or something doing it all at once
15:08<synapt>Well don't forget Linode is GMT-5
15:08<synapt>only 3PM here
15:08<Katana>one hour ahead of me, heh
15:08<synapt>so still got many hours left 'in the day' :P
15:08*Katana checks bank account, sees he has enough for that wireless adapter for the home server box
15:09<Zr40>!cloud wireless server!
15:09<linbot>I'm leaving Linode for the cloud!
15:09-!-ronkrt [] has joined #linode
15:09<Bartzy>I need to package a debian box and want to make sure it's small. What can I do besides aptitude clean ?
15:09<Bartzy>What are the obvious ways to get rid of temp/not-needed data ?
15:09<Zr40>rm -rf /
15:10<Katana>Zr40: I want something left on at home so I can VPN in :P
15:10<Katana>Zr40: also that doesn't work in latest releases, you know that right?
15:11<Bartzy>Zr40: Thanks, works rather well
15:11<Zr40>Katana: perhaps, but sudo rm -rf --no-preserve-root / doesn't quite pack the same punch
15:11<Bartzy>Please help :p
15:11-!-scorche [] has quit [Read error: Connection reset by peer]
15:11<Katana>Zr40: I know. One of these days someone will figure out how to obfuscate it so that it doesn't look as scary
15:12<Zr40>Katana: like :(){ :|:& };: ?
15:12<Bartzy>Anyone ?
15:12<Katana>Zr40: wahahahahahah
15:12<@akerl>Bartzy: Have you tried deleting what you don't need?
15:13<Bartzy>akerl: Yes. And then I thought about the apt cache. And then I came there :)
15:13<Bartzy>there is elsewhere
15:13<Zr40>Bartzy: check /tmp, /usr/tmp and /var/tmp (the last two might not exist)
15:13<@akerl>also can probably truncate some logs
15:14<@akerl>essentially, since your goal is (I assume) to use as little physical space as possible, I'd use du to see what's eating space and trim things out based on that
15:14<Bartzy>yaaaa found some stupid logs, 38MB :D
15:14<Zr40>Bartzy: also, you can use du -s * from the root directory to identify the largest directories, then cd down and repeat. You might find something big you don't need any longer
15:14<@akerl>`du -sh --exclude=/proc /*`
15:15<Zr40>yes, ignore /proc, /sys and /dev.
15:15<Bartzy>| sort -h :p
15:15<Bartzy>/usr/src/linux-headers-2.6.32-5-common , is that important to anything ?
15:16<dwfreed>Bartzy: unless you're compiling, not really
15:17<Zr40>it's needed if you need to compile anything that directly interacts with kernel functions
15:17<Zr40>it's probably installed as a package, try removing the package instead of the directory
15:17-!-ronkrt [] has quit [Ping timeout: 480 seconds]
15:18<Bartzy>Thank you
15:18<Zr40>Katana: eek. USB networking
15:18<Bartzy>BTW - anyone here is using Vagrant by any chance ?
15:19-!-scorche [] has joined #linode
15:20<Katana>Zr40: I'm not going to be using the box for anything important, so ehh
15:21<Katana>Zr40: my current problem though is I can't get into the box half the time remotely - "no route to host" even though the box has outward communication
15:22<Zr40>from where?
15:22<dwfreed>Katana: not sure how USB wifi would make that any better
15:22<Katana>Zr40: LAN.
15:22<Zr40>Katana: got an exact error handy?
15:22<Katana>Zr40: "No route to host" when I'd ping it
15:22<dwfreed>Zr40: he gave you one
15:23<Katana>that's all it'd give
15:23<dwfreed>Usually means you're getting ICMP host unreachable back
15:23<Katana>router would wtf on direct communication to it using the lan ip, but outward comms was fine from it
15:23<Katana>I'd have the monitor hooked up and be pinging the world back out, but nothing in
15:23<Katana>then for ~15 minutes it'd be fine, I'd have connection
15:23<Katana>and randomly, it'd drop out and no inbound stuff
15:24<dwfreed>Katana: listening serial modem on POTS line == old school remote desktop :)
15:24<Zr40>sounds like the router is forgetting the server exists, until the server itself tries to communicate
15:24<Zr40>as a workaround, you could set up a regular ping from the server, say, once every minute
15:24<Katana>Zr40: It'd drop the connection even when SSH'd, having trafffic back and forth
15:24<dwfreed>Cron jobs ♥
15:24<Zr40>Katana: anything relevant in the server's dmesg?
15:25<Katana>Not that I ever noticed - it seemed to be hardware level, not software
15:25<Zr40>not even link down/up messages?
15:25<Katana>ethernet connection, when I moved the box close enough and had a wire running across the hall between the box and the router, was perfectly fine
15:26<Zr40>could be the wire then
15:26<Katana>Zr40: the box never saw the link drop. the entire time it had outgoing traffic, I could browse google (when I installed a desktop distro on there), but nothing could get in
15:26<Katana>always with No route to host
15:26<Zr40>no route to host sounds like a local arp table issue
15:27<squircle>Katana: if you ssh into your router, can it communicate with this box?
15:27<Katana>Outgoing stuff was always working fine, but incoming connections for things like SSH, SFTP, HTTP, direct connections, they'd all asplode
15:27<Katana>squircle: No access to the router like that, alas. :\
15:27<squircle>awh, lame
15:28<Katana>all other systems are fine though, nothing else has that issue
15:28<Zr40>fwiw, this is what I get on osx when I ping a non-existing local address
15:28<Bartzy>Is there any easy way to check via a script if a public key is already added to ~/.ssh/authorized_keys ?
15:28<Zr40>Bartzy: grep "public key here" ~/.ssh/authorized_keys
15:29<Zr40>(don't include the key comment there)
15:29-!-teacup [] has joined #linode
15:30<Zr40>Bartzy: exit status is 0 if it exists, 1 if not
15:31<dwfreed>and you can use -q to prevent it from outputting the found line
15:32<Zr40>heh @ NOTES section in man grep
15:33<Zr40>Katana: do you happen to run a firewall on that server?
15:37-!-jhanjon [] has joined #linode
15:37-!-seanh-ansca [] has joined #linode
15:38<jhanjon>Anyone use SystemImager for copying a linode to another server?
15:38-!-advion [] has joined #linode
15:39-!-teacup [] has quit [Remote host closed the connection]
15:40-!-nisstyre [] has quit [Quit: Leaving]
15:41-!-jhanjon [] has left #linode []
15:47<nofxx>Anyone using the loadbalancer? Just so I can understand it a lil better, it's basically a webUI for haproxy?
15:47<nofxx>Or, if no haproxy, what is used?
15:47-!-seanh-ansca [] has quit [Quit: Leaving.]
15:48<Katana>Zr40: Nein. used a stock ubuntu image.
15:48<niftylettuce>trying to proxy nging/node
15:49<niftylettuce>nginx* -- do I need to set `root` for the config for the assets to get served properly?
15:49<niftylettuce>so the assets get served properly*
15:54-!-nisstyre [] has joined #linode
15:56<Zr40>Katana: tried watching tcpdump when the error happens?
15:58-!-seanh-ansca [] has joined #linode
16:08-!-seanh-ansca [] has quit [Quit: Leaving.]
16:17<Zr40>niftylettuce: root controls the base path where files are served
16:18<niftylettuce>Zr40: yea i need to use app middleware to get this to work
16:18<niftylettuce>since im using which then gets routed to a port/node
16:18<niftylettuce>so when i do I get a 404 on /somepath when it should be loading the index
16:18<Zr40>niftylettuce: similar here. I'll show you my config
16:18<niftylettuce>Zr40: cool please share
16:19<Zr40>actually noticed my root is wrong. *goes and fixes*
16:20<Zr40>anyway, in my case, it should point to the correct public directory of my app
16:24<niftylettuce>Zr40: but you're using a subdomain, im not
16:24<niftylettuce>i need to do a rewrite i think
16:24<Zr40>niftylettuce: does it matter? just change server_name
16:26<niftylettuce>Zr40: you're running node on port 80?
16:26<Zr40>nginx on 80, unicorn on a unix socket
16:26-!-synapt [] has quit [Ping timeout: 480 seconds]
16:27-!-synapt [] has joined #linode
16:27<Zr40>if node can't listen on a unix socket, just change the upstream server to instead of unix:/path/to/socket
16:28-!-AviMarcus [~avi@] has joined #linode
16:30<Zr40>actually, just No http://
16:32<niftylettuce>Zr40: hm, but it looks like that would rewrite all requests
16:32<niftylettuce>Zr40: I still want to be able to serve up static files in
16:33<Zr40>niftylettuce: an upstream block doesn't do anything by itself
16:33<Zr40>niftylettuce: in my config, see the try_files line
16:34<Zr40>niftylettuce: it first tries to match the request against a file in my app's public directory
16:34<Zr40>if it doesn't exist, it forwards it to @app, which is defined in the location block just below
16:34<Zr40>that block proxies the request to http://steamstats, which is the upstream defined above
16:34<niftylettuce>ok let me try!
16:36<Zr40>technically you don't need the upstream block if there's only a single backend server
16:36<Zr40>but it allows you to add more without changing the config too much
16:36-!-synapt [] has quit [Ping timeout: 480 seconds]
16:40-!-luisvelaz [] has joined #linode
16:40-!-luisvelaz [] has quit [Remote host closed the connection]
16:45-!-goose [] has quit [Quit: Staying alive can kill you, it's taken years off of my life.]
16:46<niftylettuce>Zr40: im just getting a bunch of 404's
16:47<niftylettuce>Zr40: do I need that rewrite?
16:48<Zr40>the rewrite I have at the bottom? no, that's just a redirect for an old URL
16:49<Zr40>niftylettuce: does /srv/www/blah/public contain the files you want?
16:50<niftylettuce>Zr40: yeah
16:51<Zr40>are they readable by the user nginx is running under?
16:51<niftylettuce>Zr40: yeah nginx running as sudo
16:51<@heckman>interesting user
16:51<niftylettuce>er main user ubuntu
16:51<Zr40>nginx drops privileges to the user you specify using the user line (line 1 in my config)
16:52<Zr40>if it's not specified, it probably defaults to 'nobody'
16:52<niftylettuce>user deployer users; ?
16:52<niftylettuce>so deployer is a user?
16:52<niftylettuce>and users is a group?
16:52<Zr40>in my case it is, to both
16:53<Zr40>it should be a user that can read the files and has no additional privileges
16:54<niftylettuce>Zr40: yeah they are readable by user nginx is running under
16:54<Zr40>niftylettuce: which is?
16:56<niftylettuce>Zr40: www user
16:56<niftylettuce>and i have that at the top of the file
16:57<Zr40>okay, try this: sudo -u www cat /srv/www/blah/public/<request path which should work but results in a 404>
16:57-!-jbw [] has quit [Ping timeout: 480 seconds]
16:58<@heckman>Wait, 'www'? That's not a default Ubuntu user
16:59-!-synapt [] has joined #linode
17:00<niftylettuce>heckman: no worry, i have it setup right
17:00<niftylettuce>Zr40: the cat output was right
17:00<Zr40>niftylettuce: so... http://blah/<exact same request path> doesn't work?
17:02<niftylettuce>Zr40: nope
17:02<@akerl>niftylettuce: Paste the exact error from your logs?
17:02<@akerl>(in the pastebin, not here)
17:02<niftylettuce>im not that big of a noob :D
17:03<niftylettuce>GET /blah/ 404
17:03<niftylettuce>that is repeated like 500 times for all my request attempts
17:03<@akerl>niftylettuce: Paste the exact error from your logs?
17:04<@ericoc>akerl: !
17:04<Zr40>niftylettuce: I think you didn't use the exact same path for the cat and the http request
17:06-!-jbw [] has joined #linode
17:06<niftylettuce>I used the _exact_ same path
17:06<@akerl>that's cool.
17:06<Zr40>but your log paste shows you accessed a directory, which cannot be cat
17:07<niftylettuce>Zr40: it is not getting routed through properly
17:08<Zr40>niftylettuce: are you supposed to get a static file for that request or one from the backend server?
17:08<niftylettuce>Zr40: static index.html
17:08<Zr40>change your try_files $uri @app; to try_files $uri $uri/index.html @app;
17:09<niftylettuce>well actually, it should be dynamic since I could add routes in node
17:09<niftylettuce>i mean, I use express.static server to serve up assets in the /public folder
17:09<@akerl>if you have a proper index configured, you should just be able to use $uri $uri/ @app
17:09<Zr40>which would be
17:11<Zr40>so, for the backend
17:11<Zr40>do requests to a path you know is handled by node work?
17:12<niftylettuce>let me try that
17:13<niftylettuce>Zr40: no, doesn't, I added app.get('/test', ...); and tried and got 404
17:14<Zr40>does /test without blah work?
17:16<niftylettuce>Zr40: no
17:17<niftylettuce>would a rewrite be easier than all this?
17:17<Zr40>on the server, does curl http://localhost:1337/test work?
17:17<niftylettuce>Zr40: aha!
17:17<niftylettuce>Zr40: so heres whats up, I don't even get a new entry in access.log when I load
17:17<niftylettuce>which means that was from before when I was fooling around, so its not routing properly
17:18<Zr40>node's or nginx's access log?
17:18<niftylettuce>Zr40: nginx
17:18<niftylettuce>curl locally works
17:20<Zr40>so, if I understand you correctly
17:20<Zr40>you want to point to node's /test
17:20<niftylettuce>Zr40: yup
17:21-!-Ghost [~Ghost@] has joined #linode
17:21<Zr40>then you need to change proxy_redirect off; to proxy_redirect http://blah/ /blah/; (with the first blah being the name of the upstream block)
17:21<niftylettuce>location @app should be changed?
17:21-!-Ghost [~Ghost@] has quit []
17:23<Zr40>this change removes /blah before proxying the request to node
17:24<Zr40>...I think.
17:24<Zr40>the docs for that directive are a bit vague
17:24<gdr>fantastic morning music
17:24-!-vivid_ac_ [] has joined #linode
17:24<Zr40>anyway, I should be going to bed. Good luck, niftylettuce :)
17:25<niftylettuce>Zr40: thanks, i dont think that worked though
17:25<niftylettuce>Zr40: but gnite and tyvm for all your help
17:25-!-zivester [] has quit [Read error: Connection reset by peer]
17:25-!-zivester [] has joined #linode
17:26-!-AviMarcus [~avi@] has quit [Quit: Ex-Chat]
17:28-!-Bartzy [~bar@] has quit [Ping timeout: 480 seconds]
17:30-!-vivid_ac [] has quit [Ping timeout: 480 seconds]
17:32<niftylettuce>akerl: any advice?
17:36<@akerl>niftylettuce: Your config is borked, or your permissions are borked
17:43-!-dubenstein [~dubenstei@] has quit [Ping timeout: 480 seconds]
17:53-!-nmudgal [~tracker@] has quit [Ping timeout: 480 seconds]
17:57-!-Bryen [~bryen@] has quit [Read error: Connection reset by peer]
18:03<numk>niftylettuce: what's your proble?
18:03-!-BarkariII [BarkerJr@2002:1880:41b:1:982d:6d4e:5a8a:cdd7] has quit [Remote host closed the connection]
18:03-!-BarkerJr [BarkerJr@2002:1880:41b:1:982d:6d4e:5a8a:cdd7] has joined #linode
18:09-!-ronkrt [] has joined #linode
18:11-!-Bartzy [] has joined #linode
18:17<niftylettuce>numk: using express.js/node.js and trying to proxy to it on path, like -- whereas it points to which in turn loads an express.static server from /srv/blah/public
18:21<niftylettuce>oh shit
18:21<niftylettuce>the reason why it wasnt working is because ec2 blocks all ports other than 80
18:21<niftylettuce>err maybe not
18:21<niftylettuce>MAYBE just maybe, im unsure though i have to check
18:21<niftylettuce>but thts what im thinking
18:22<bob2>put the bong down
18:22<bob2>proxy to localhost:1234
18:22<bob2>if you need to strip the path, rewrite
18:22<niftylettuce>yea wouldnt be ec2 since i can curl
18:22<niftylettuce>and its local
18:22<bob2>it's not ec2, it's your config
18:23<niftylettuce>f the config >:
18:23-!-ronkrt [] has quit [Read error: Connection reset by peer]
18:23<bob2>ok but the problem is with your config
18:23-!-ronkrt [] has joined #linode
18:23<niftylettuce>yeah i have no clue how to fix this
18:23<niftylettuce>all tutorials are misleading/different cases
18:23<bob2>i doubt that
18:29-!-nicinabox [] has joined #linode
18:30-!-saikat [~saikat@] has joined #linode
18:32-!-saikat [~saikat@] has quit []
18:39<nicinabox>i need some help debugging a high cpu issue with mysql
18:40-!-robzilla [] has quit [Read error: Connection reset by peer]
18:40<linbot>If you have a question, feel free to just ask it -- someone's always willing to help. If you don't get a response right away, be patient!
18:40<nicinabox>wasn't sure if anyone was around :P
18:41<nicinabox>i've got a standalone mysql server
18:41<nicinabox>i frequently get alerts for > 90% cpu
18:41<nicinabox>i've narrowed it down to the application that's likely causing the spikes
18:41-!-robzilla [] has joined #linode
18:42<nicinabox>the app will make the cpu just to 100% instantly, but there's almost no load
18:43-!-Ricki [~Ricki@] has joined #linode
18:43<Kyhwana>Well, you need to figure out how to profile that app then
18:43<nicinabox>here's a shot from htop when it's connected
18:44<nicinabox>is this something i should be concerned about?
18:44<niftylettuce>hm im getting ERR_CONN_REFUSED with this
18:45<Kyhwana>I dunno, is your app really that DB intensive?
18:45<nicinabox>it's not. or at least it doesn't seem to be
18:45<Kyhwana>nicinabox: because you don't own ?
18:45<nicinabox>they're one of our clients
18:46<Kyhwana>er, niftylettuce rather
18:46<nicinabox>it's basically just a cms
18:46<bob2>show processlist
18:46<Kyhwana>nicinabox: well, unless the app is going slow, I wouldn't worry about it
18:46<niftylettuce>Kyhwana: no lol thats not it
18:46<nicinabox>Kyhwana: well, they actually have reported slowness. i'm wondering if that's from the db?
18:51<Kyhwana>nicinabox: possibly, or your app sucks
18:51<nicinabox>Kyhwana: ha, i have no doubts about that.
18:52<bob2>show processlist
18:53<Kyhwana>do what bob2 said
18:53<bob2>also turn on slow query log
18:53<bob2>also us the db profiling stuffyour app no doubt has because not havbing that would be stupid
18:54<nicinabox>bob2: i turned on slow query log a few days ago
18:55<nicinabox>we aren't the maintainers of said app though
18:55<bob2>that's fine
18:55<bob2>you're the sysadmin though
18:56-!-DephNet[Paul] [] has quit [Ping timeout: 480 seconds]
18:57<niftylettuce>bob2: what should my proxy_redirect be?
18:58<niftylettuce>or should I use a rewrite?
18:59<joshie_>w00t using redhat
18:59<@heckman>I'm sorry for your loss
18:59<niftylettuce>ah i think i got it
19:01<nicinabox>what can i do to lower the strain on the db?
19:02<Kyhwana>nicinabox: make less queries?
19:02<nicinabox>would varnish work?
19:02<joshie_>i like amazonaws >.>
19:03<praetorian>don't read static content from a db.
19:04<@heckman>nicinabox: you can give your database more RAM to cache more stuff, use memcached, etc. You could also cache entire pages too, using something like varnish
19:15-!-thews [] has quit [Quit: thews]
19:15-!-ronkrt [] has quit [Read error: Operation timed out]
19:16-!-Boohemian_ [] has joined #linode
19:17-!-Ricki [~Ricki@] has quit [Ping timeout: 480 seconds]
19:19-!-Boohemian [] has quit [Ping timeout: 480 seconds]
19:20-!-seanh-ansca [] has joined #linode
19:20<Kyhwana> time to update your firewalls
19:22<EugeneKay>I use the "evil bit"
19:22<Kyhwana>EugeneKay: NULL packets are Good, since they can't have the Evil Bit set
19:23<niftylettuce>how can I do a proxy_cache so that each time the page loads, it does not load it from cache, aka its a fresh load
19:23<niftylettuce>found this but not sure which to use
19:23<EugeneKay>Not to sound obvious, but have you looked for something without the _cache?
19:24<niftylettuce>proxy_cache no-cache ?
19:24-!-seanh-ansca [] has quit []
19:24<Kyhwana>See NULL bits above
19:30<linbot>New news from forums: DB-serverless blog/CMS software suggestion? in Web Servers and Web App Development <>
19:32*heckman goes back to fighting with Python/M2Crypto
19:34<Kyhwana>heckman: had a look at libtomcrypt? I think someones written python bindings for it
19:35<@heckman>I'm using the OpenSSL stuffs in M2Crypto
19:35<@heckman>Specifically for HTTPS
19:35<Kyhwana>ahh right
19:36<praetorian>look what happesn when you have a 'secret' game sale
19:36<@heckman>I've got it working if I only connect to something *once*
19:36<@heckman>If I try multiple connections, it excepts claiming it got no cert.
19:37<@heckman>It's lying...
19:38<@heckman>I wish there was an easy way to force a module to get bent in Python and re-init.
19:39-!-danblack [] has joined #linode
19:42-!-synapt [] has quit [Ping timeout: 480 seconds]
19:42-!-ronkrt [] has joined #linode
19:43-!-robzilla [] has quit [Read error: Connection reset by peer]
19:44-!-Bryen [~bryen@] has joined #linode
19:47<Katana>why in the world would someone want to run a web app off of sqlite
19:48<Katana>jesus christ you're just asking for file lock issues
20:01-!-Boohemian [] has joined #linode
20:01<Kyhwana>probe dagmem dsize=512M nid=0:1:1:1:1 64bit=1:1:1:1:1
20:01<Kyhwana>FATAL: Error inserting dagmem (/lib/modules/2.6.32-220.7.1.el6.x86_64/extra/dagmem.ko): Unknown symbol in module, or unknown parameter (see dmesg)
20:01<Kyhwana>ignore that -.-
20:02<niftylettuce>hey guys, if I want to redirect to just when using a proxy to port 1234, is this the proper rewrite? `rewrite ^blah/(.*)$ /$1 break;`
20:04-!-Boohemian_ [] has quit [Ping timeout: 480 seconds]
20:06-!-wheatie [] has joined #linode
20:06-!-azaghal [~azaghal@] has quit [Remote host closed the connection]
20:07-!-azaghal [~azaghal@] has joined #linode
20:09-!-danblack [] has quit [Ping timeout: 480 seconds]
20:14-!-Bartzy [] has quit [Ping timeout: 480 seconds]
20:21<niftylettuce>im out, nginx ftl
20:25<fayimora>would a 512 be ok for 1.5million users on the same day?
20:26<pharaun>context? what you running? etc?
20:26<pharaun>static site? probably, php_mod, ha ha ha
20:27<fayimora>rails site
20:28<fayimora>just a few post requests
20:28<fayimora>login, fetch data and display.. dats all
20:29<pharaun>caching, you has it
20:30-!-pjkh [] has quit [Quit: Leaving.]
20:31-!-fayimora [] has quit [Quit: Busy…..zzzzz]
20:32-!-fayimora [] has joined #linode
20:34-!-hipsterslapfight [] has quit []
20:34<fayimora>pharaun: sorry i missed ur reply
20:34<fayimora>got disconnected
20:34<pharaun>caching, you has it?
20:35<fayimora>I don't get
20:35<fayimora>so what do i cache? An would a 512 be sufficient?
20:37<fayimora>ok so bcd to the question.. would a 512 be sufficient after caching?
20:38<@heckman>fayimora: it's almost impossible to say without implementing it.
20:38<fayimora>hmm ok thanks
20:38<joshie_>try and find out!
20:38<joshie_>sometime in the near future.. lincloud
20:38<linbot>I'm leaving Linode for the cloud!
20:44-!-Kabaka [] has quit [Quit: BRB]
20:45-!-Kabaka [] has joined #linode
20:46<squircle>I think the clock on this server is broken... every minute ntpd adjusts the time back by 1.5-2 seconds
20:46<squircle>i was about to file a support ticket until i remembered it isn't a linode
20:47-!-nofxx [] has quit [Remote host closed the connection]
20:53-!-Nemykal_ [] has joined #linode
20:55-!-Nemykal [] has quit [Read error: Operation timed out]
20:55-!-mkasu is now known as maku`off
20:56-!-SnakeO [] has joined #linode
20:57-!-Dedalo [~Dedalo@] has quit [Remote host closed the connection]
21:13-!-BlendedByChris [] has joined #linode
21:14-!-mdcollins [] has joined #linode
21:14-!-nicinabox [] has quit [Remote host closed the connection]
21:15-!-DreamPhysix [] has joined #linode
21:20-!-radhermit [~radhermit@2600:3c01::f03c:91ff:fe96:db5a] has quit [Ping timeout: 480 seconds]
21:24-!-DreamPhysix [] has quit [Ping timeout: 480 seconds]
21:24-!-TeddyR [] has quit []
21:28-!-ido [] has quit [Quit: leaving]
21:28-!-ido [] has joined #linode
21:33-!-thews [] has joined #linode
21:37-!-SelfishMan [] has quit [Quit: meh]
21:43-!-radhermit [~radhermit@2600:3c01::f03c:91ff:fe96:db5a] has joined #linode
21:46-!-Boohemian [] has quit [Ping timeout: 480 seconds]
21:49<Katana>"why aren't you making them do things more better"
21:56-!-atomaka [] has joined #linode
21:57-!-radhermit [~radhermit@2600:3c01::f03c:91ff:fe96:db5a] has quit [Ping timeout: 480 seconds]
21:58-!-gadams [~IAmMrAwes@] has joined #linode
22:02-!-tolle [] has joined #linode
22:05<squircle>this chinese ISP's response to an abuse email from my SSH honeypot was just instructions to change the sshd port
22:05<squircle>"our users aren't abusive, you're just doing it wrong"
22:06<squircle>meanwhile it's being wordlist-username-attacked by this one IP (amanda, alex, alexander.... robin, robert, root, r00t...)
22:06<gadams>I enabled ufw and now I can't apt-get update/upgrade, bundles of fun.
22:06<Kyhwana>Which honeypot are you running?
22:07<squircle>a modified version of kojoney
22:08<squircle>although I may write my own as a little fun project
22:09<Kyhwana>Wasn't there a forked/updated version of that?
22:10<Kyhwana>oh, kippo
22:10<squircle>probably; I set this thing up years ago and haven't touched it
22:10<Kyhwana>oh, it was inspired by, not based on
22:10<squircle>I need to change it though; I have to add more hack-y code to it every day
22:11<squircle>yeah, I saw that
22:11*Kyhwana ponders getting a cheapo VM elsewhere to run one on
22:11<squircle>hah... "Daily abuse report... Login attempts by RIR: APNIC: 1.1k, ARIN: 27..."
22:11-!-Linear [] has joined #linode
22:12<Kyhwana>squircle: does yours run on v6?
22:12*squircle checks
22:12<squircle>it would appear so, yes
22:13<Kyhwana>Would be interesting to see how many v6 brute forcers there are around
22:13<squircle>i'll see if I can add a hook for it
22:13<squircle>use ALL my free time
22:13<Kyhwana>probably not many, since it'd take forever to try all the IPs :P
22:14<squircle>they make python for AS/400?!?!
22:14<Kyhwana>I moved my ssh port and fired up netcat on port 22 v6 only and didn't get anything :P
22:14<squircle>yeah, I wouldn't expect too many
22:14<squircle>unless they get it from a list somewhere
22:16<avenj>yet to see any kind of v6 abuse
22:17<Kyhwana>squircle: had anyone (D)OS your ssh honeypot host?
22:21-!-advion [] has quit [Quit: This computer has gone to sleep]
22:22<gadams>Anyone ever turn ufw on and have apt-get update/upgrade fail on them because packets aren't being received ?
22:23<Kyhwana>gadams: uh, so you need to change your rules to allow that?
22:23-!-SelfishMan [] has joined #linode
22:23<Kyhwana>squircle: DDOS, even
22:24<gadams>Kyhwana, I've got 22,80,443 to allow * am I missing some other magical port?
22:25<Kyhwana>what port are the packets coming back into?
22:26<gadams>Good question.
22:26<gadams>Let me find out.
22:27<squircle>Kyhwana: occasionally, yes
22:27<squircle>Kyhwana: but my NSP is very proactive in mitigating it
22:27<Kyhwana>squircle: hmm, that's why I don't wanna run it on my linode :P
22:28<squircle>Kyhwana: yeah, not a great idea. my honeypot has the ability to (essentially) tell this company's core routers to discard all traffic from a certain IP or subnet
22:28<squircle>(inbound bandwidth isn't a problem, unless it's a massive global attack)
22:28<squircle>they've got 10 or 12 gigabits/s worth, so... :P
22:34*gadams is an id10t
22:34-!-TimTim` [] has joined #linode
22:34-!-TimTim` [] has quit []
22:37-!-fayimora [] has quit [Quit: Busy…..zzzzz]
22:39<auraka>no linode april sad....I figured they'd at least announce jed was coming back or that that the restroom outside caker's office was now for community use
22:44-!-foreverwondering [] has joined #linode
22:44-!-foreverwondering [] has left #linode []
22:48-!-piedlourde [] has joined #linode
22:49<Gshock>Whats the simplest way to send email from multiple vhosts? I have EXIM set up based on the guide, but I dont think it works quite right for multiple domains
22:53-!-TimTim [] has quit []
22:56<Kyhwana>how are vhosts involved in sending email?
22:56<Kyhwana>auraka: or that employeers are required to pee in the planets, because watering them is too expensive
22:57<Gshock>Kyhwana: thats part of the issue I'm not even certain I'm asking the right question.
22:57<auraka>Kyhwana: with some of the talent I'm sure that is happening already
22:59<Kyhwana>auraka: ahaha
22:59<Kyhwana>wonder where dwfreed is
22:59<auraka>"watering plants"
22:59-!-teacup [] has joined #linode
22:59<Kyhwana>Gshock: you mean you want to use your linode to send email from more than one domain?
22:59<linbot>New news from forums: Is anyone else getting rebooted every Sunday at 6:25am? in General Discussion <>
23:00<Kyhwana>^ wat
23:00<linbot>I want a free hat!
23:00<Gshock>Yes. Its properly sending, but its getting marked as spam in some instances I think. It shows up as coming from my linode, with no reverse DNS (which I'm sure is the problem) but I'm not quite sure how to move forward
23:01<Gshock>I'm using EXIM right now, but I'm open to suggestion. EXIM asks for your systems FQDN and thats where I put my linode info.
23:01<Kyhwana>Gshock: so set the PTR record for your linodes IP to be what you want it to be?
23:01<Kyhwana>what was the FQDN you put in?
23:02-!-Kabaka [] has quit [Quit: s/Kabaka//]
23:02<Gshock>Which I'm sure is probably wrong.
23:02<Kyhwana>you want the actual domain your hosting email for?
23:02<Gshock>Right, so do I enter all the domains?
23:02-!-Kabaka [] has joined #linode
23:03<Gshock>I'm hosting multiple sites.
23:04<Kyhwana>I think (but im not 100% sure) that you want something like "" and "" resolves to PTR for is and both resolve to
23:04<Kyhwana>Also setup SPF and DKIM for all the domains your hosting mail for. (vis DNS records)
23:05-!-Kunda [] has joined #linode
23:05<linbot>New news from forums: Dashboard feature request in Feature Request/Bug Report <>
23:05<Gshock>I've seen SPF and DKIM a few times, what is "PTR" in this context?
23:05<Kyhwana>So you'd put in "" as your FQD and make sure that it resolves to your linodes IP and that your linodes IP reverse resolves (PTR) to
23:06<Kyhwana>PTR is pointer record. That's what your IP reverse resolves to. It's a type of DNS record
23:06<SleePy>Curse you DirectTV and your problems!
23:06<Gshock>Do you have any setups that send email from multiple domains? Curious if I shouldnt be using something other than EXIM.
23:06<Kyhwana>nope, someone else might
23:06<Kyhwana>(I just use google apps)
23:06<bdube>Gshock: I use Postfix and send from multiple names
23:07<Kyhwana>But first of all, do what i've said above.
23:07<Gshock>bdube: would you happen to know of a good tutorial for it?
23:07<Kyhwana>You definetly need/want the SPF, DKIM and DNS issues fixed
23:08<bdube>Gshock: A combination of a few Linode Library guides will get you almost all the way there
23:08-!-teacup [] has left #linode []
23:09-!-ZeeO [] has quit [Read error: Connection reset by peer]
23:09<Gshock>Yeah, the guides are just written like they are for someone that already understands what they are doing and therefore wouldnt need the guide.
23:10<bdube>Gshock: Remember that your VPS has only one canonical name, no matter what/how many things you host on it. Get your A/AAAA/PTR records straight for that, then add your multiple service scenario.
23:10<Gshock>As an example, you need a "system mail name" which is a FQDN, but I'm unclear what actually goes there, certainly its not, and almost as certainly its not a random clients domain
23:10<Kyhwana>Gshock: are you using the linode DNS servers to do your DNS?
23:11-!-Kunda [] has quit [Quit: Kunda]
23:11<gadams>Kyhwana: Alright, 99% positive it isn't the firewall with apt-get not connecting to anything. No logs in the firewall either about it denying connections. Any ideas?
23:11<Kyhwana>gadams: turn debug mode in ufw on?
23:11<Kyhwana>what port on the server does apt-get connect to? ;)
23:12<Kyhwana>and then, whats the source port for apt-get? You're not blocking all incoming packets are you?
23:13*heckman shakes fists
23:13<Kyhwana>Note that iptables has "-m conntrack --ctstate ESTABLISHED,RELATED". Surely UFW has something similar?
23:14*gadams I HATE FIREWALLS
23:14<Turl>for the record, ufw doesn't block outgoing connections by default
23:15<Kyhwana>Turl: so it allows incoming packets from established connections?
23:15<Kyhwana>gadams: welp, you probably did it wrong then ;)
23:16<gadams>I even disable ufw and still get an err 113 from apt-get update/upgrade
23:17<Kyhwana>whats error 113?
23:17<bdube>Gshock: The canonical name goes there. Right, likely not *, likely not a random client's domain. It's arbitrary in a sense, but still important.
23:17<Gshock>Thats what I'm not getting.
23:17<Turl>gadams: the mirror you're using is just broken?
23:17<Kyhwana>Gshock: use your own domain that you've got pointed at your linode?
23:17<Turl>maybe the*
23:17<Gshock>arbitrary as in I pick whatever I want and shove it in there?
23:17<bdube>Gshock: Not now, no. It was arbitrary when you would have chosen the hostname.
23:18<Gshock>I currently don't have such a domain. Long story short my boss did a bunch of really stupid things with our company website and I cant move that site just yet
23:19<Gshock>When I chose the hostname for what, when I was configuring the email server? Or when I was purchasing a domain/
23:19<bdube>Neither, really
23:19<Gshock>I can reconfigure it to say whatever I want.
23:19<gadams>Turl: all of them?
23:19<Kyhwana>Gshock: if it's a FQDN, it has to resolve.. ie
23:19<Turl>gadams: working fine here, is that one failing on you?
23:20<Gshock>So, basically i need to use an FQDN that I want to represent every domain I'm hosting?
23:21<gadams>Turl, running ubuntu :-|
23:21<Gshock>so someone gets an email from, and it says "via"
23:21-!-Boss [] has joined #linode
23:21<Turl>gadams: well, what mirror did you try? I can check from here
23:21<bdube>Gshock: No, you won't have that via crap when it's configured correctly
23:22<linbot>Katana: IP:; rDNS: None; City: Westlake Village; State: California; Postal code: 91361; Country: United States; Domains: 8;
23:24<Gshock>Right, its pretty easy to google info on that
23:25<Kyhwana>DNS wise, what I said early applies. You need a "primary" domain for your IP to reverse/resolve to (
23:25<Gshock>Making EXIM or Postfix do that, now that appears to be a different thing entirely.
23:25<Kyhwana>Gshock: yes, you need to have users auth as ""
23:25<Kyhwana>there must be something that tells you how to do that
23:26<Gshock>When you say I need to have users "auth"... just curious - I dont want to receive email, just send only.
23:26<Gshock>This guide looks like its massively more than I should need. Maybe I'm wrong.
23:26<Gshock>This guide.. doesnt quite seem to get to where I need
23:27<Kyhwana>Gshock: erm yes. You still want users to prove they are who they say they are
23:27<Kyhwana>Otherwise the janitor can send email through you as "", which would be bad, right?
23:27-!-Jaimechen [] has joined #linode
23:27<Gshock>I'm the only user. I'm hosting Drupal sites which need to be able to send an occasional email from a form or to a user for registration.
23:29<Kyhwana>OK, so you don't have anyone else connecting to your server to send email through you as I said above? Then just go ahead and do all those DNSey things as per above and find a tutorial that tells you how to send email via a form/etc
23:30<Gshock>Drupal is a CMS that handles that, as long as the server is configured to send mail.
23:30<bdube>Kyhwana: If the canonical name is, I would set the PTR to and not *shrug*
23:31<Kyhwana>bdube: *shrug* it was an example. as you said, it doesn't matter as long as it resolves to the same thing both ways
23:32<gadams>turl: 0% [Connecting to (] [Connecting to (] [Connecting to (]
23:32<bdube>Kyhwana: Fair enough. I wasn't sure if you meant to set the PTR to mail.blah.blah no matter what, because that machine handles mail
23:33-!-pygmalio1 [] has quit [Quit: leaving]
23:33<Turl>gadams: looks operational from here
23:33<Turl>gadams: routing issue?
23:34<Katana>gadams: firewall in the way? checked your logs?
23:35-!-tharkun [~0@] has quit [Ping timeout: 480 seconds]
23:35<Kyhwana>disable the firewall?
23:35<gadams>Katana: I had enabled UFW now disabled, still getting RIP'd. with apt-get update on 10.04
23:35<gadams>heckman, you working tonight ?
23:36<Katana>gadams: tail /var/log/ufw.log, double-check.
23:36<Turl>gadams: can you wget any other random website?
23:36<Katana>Turl: apt isn't over http
23:36-!-Jaimechen [] has quit [Remote host closed the connection]
23:37<@heckman>gadams: nope
23:37<purrdeta>w 30
23:37<@heckman>I am doing Tues - Sat 8a-5pmish now
23:37<purrdeta>so sorry I am dumb
23:37<Katana>heckman: can you teach me the recipe to your secret bbq sauce
23:37<praetorian>i went and purchased a stylus ($2.50) and it doesnt work with draw something!
23:37<praetorian>how am i to ever draw non stick people!
23:38<Katana>praetorian: with your TEETH
23:38<praetorian>good point
23:38<Turl>Katana: it is
23:38<praetorian>heckman: do you draw something? perizomg hasnt drawn me back >:(
23:39<gadams>Katana: no IPs showing up
23:39<gadams>turl: yes
23:40<Turl>gadams: try some other mirror then, like
23:40<@heckman>^ those go to the same plae
23:40<@heckman>US/UK Ubuntu archives are actually in the UK
23:41<Turl>heckman: I got two diff IPs from here heh
23:41<@heckman>afaik they are both in the UK, tho
23:42<Turl>uk is
23:42<Turl>us is a huge round robin, all of 91.189.x.x
23:42<Turl>more precisely, 91.189.92.x
23:43<linbot>Turl: IP:; rDNS: None; Country: United Kingdom;
23:43<linbot>Turl: IP:; rDNS: None; Country: United Kingdom; Domains: 3;
23:44<Turl>well try with then
23:49<Kyhwana><x> openvz VPS plans for $15/year. heh. *fires up ssh honeypot on one of those*
23:51<@heckman>Do you just parse logs using a script for that?
23:51-!-ronkrt [] has quit [Ping timeout: 480 seconds]
23:55<Turl>Kyhwana: I've seen those :)
23:55<Turl>I heard they're actually decent for the price heh
23:57<KyleXY>Depends on what you're running
23:57<KyleXY>for an SSH Honeypot, sure.
23:57<Turl>yeah, they're 256M boxes with 10G storage tops or such
23:58<Boss>into ubuntu eh @ heckman
23:58<Turl>for running an irc bouncer, public file hotlinking or the like they're more than enough
23:59-!-VS_ChanLog [] has left #linode [Rotating Logs]
23:59-!-VS_ChanLog [] has joined #linode
23:59<linbot>Point (0.32644015, 0.37249155) lies within the unit circle. Hits: 110535 of 140314 (π ≈ 3.151075445073193 - 0.009482791483400). http://π
23:59<@heckman>Boss: what makes you say that? :p
23:59<KyleXY>linbot gotten a lot faster lately
23:59*Boss wonders if heckman is also a gentooer
---Logclosed Mon Apr 02 00:00:28 2012