[00:02:58] RECOVERY - Puppetmaster HTTPS on stafford is OK: HTTP OK HTTP/1.1 400 Bad Request - 336 bytes in 4.912 seconds [00:05:17] New patchset: Asher; "* for mysqlatfacebook 5.1 r3875 * moved packages to unique names and /usr/local/ to co-exist with mysql 5.5 packages in ubuntu precise (some of which we bring in as dependencies)" [operations/debs/mysqlatfacebook] (master) - https://gerrit.wikimedia.org/r/15989 [00:06:57] notpeter: want to review ^^ with your new fangled dba hat? [00:26:44] +17168, -0 [00:26:45] lol [00:37:46] PROBLEM - Puppetmaster HTTPS on stafford is CRITICAL: CRITICAL - Socket timeout after 10 seconds [00:38:31] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [00:44:04] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.39 ms [00:46:37] RECOVERY - Puppetmaster HTTPS on stafford is OK: HTTP OK HTTP/1.1 400 Bad Request - 336 bytes in 2.660 seconds [00:49:01] PROBLEM - Host mw1074 is DOWN: PING CRITICAL - Packet loss = 100% [01:02:40] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [01:08:13] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.51 ms [01:12:25] PROBLEM - Puppet freshness on virt1002 is CRITICAL: Puppet has not run in the last 10 hours [01:15:25] PROBLEM - Puppet freshness on virt1001 is CRITICAL: Puppet has not run in the last 10 hours [01:21:25] PROBLEM - Puppet freshness on maerlant is CRITICAL: Puppet has not run in the last 10 hours [01:21:34] PROBLEM - Puppetmaster HTTPS on stafford is CRITICAL: CRITICAL - Socket timeout after 10 seconds [01:31:27] New patchset: awjrichards; "Enable mobile redirects for all *.wikimedia.org sites, except commons" [operations/debs/squid] (master) - https://gerrit.wikimedia.org/r/16000 [01:31:55] RECOVERY - Puppetmaster HTTPS on stafford is OK: HTTP OK HTTP/1.1 400 Bad Request - 336 bytes in 0.029 seconds [01:36:33] New patchset: awjrichards; "Enable mobile redirects for all *.wikimedia.org sites, except commons" [operations/debs/squid] (master) - https://gerrit.wikimedia.org/r/16000 [01:41:58] PROBLEM - MySQL Slave Delay on db1025 is CRITICAL: CRIT replication delay 250 seconds [01:42:52] PROBLEM - MySQL Slave Delay on storage3 is CRITICAL: CRIT replication delay 272 seconds [01:47:39] New patchset: awjrichards; "Enable mobile redirects for all *.wikimedia.org sites, except commons and bits. Change-Id: I51956cb80f6b7fa4eb9fe8d4e26047cf1181ba35" [operations/debs/squid] (master) - https://gerrit.wikimedia.org/r/16000 [01:49:46] PROBLEM - Misc_Db_Lag on storage3 is CRITICAL: CHECK MySQL REPLICATION - lag - CRITICAL - Seconds_Behind_Master : 684s [01:50:58] RECOVERY - MySQL Slave Delay on db1025 is OK: OK replication delay 11 seconds [01:54:34] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [01:54:52] RECOVERY - MySQL Slave Delay on storage3 is OK: OK replication delay 5 seconds [01:55:46] RECOVERY - Misc_Db_Lag on storage3 is OK: CHECK MySQL REPLICATION - lag - OK - Seconds_Behind_Master : 20s [02:00:07] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.39 ms [02:04:01] PROBLEM - Puppetmaster HTTPS on stafford is CRITICAL: CRITICAL - Socket timeout after 10 seconds [02:12:52] RECOVERY - Puppetmaster HTTPS on stafford is OK: HTTP OK HTTP/1.1 400 Bad Request - 336 bytes in 7.170 seconds [02:20:22] PROBLEM - Puppet freshness on db29 is CRITICAL: Puppet has not run in the last 10 hours [02:47:22] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [02:52:55] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.37 ms [03:09:53] RECOVERY - Puppet freshness on zinc is OK: puppet ran at Thu Jul 19 03:09:46 UTC 2012 [03:11:22] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [03:16:55] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.41 ms [03:17:59] New patchset: Aaron Schulz; "Added multiwrite backend config - not yet used." [operations/mediawiki-config] (master) - https://gerrit.wikimedia.org/r/15963 [03:25:10] PROBLEM - Host search32 is DOWN: PING CRITICAL - Packet loss = 100% [03:27:07] RECOVERY - Host search32 is UP: PING OK - Packet loss = 0%, RTA = 0.24 ms [03:33:25] PROBLEM - Puppet freshness on ms3 is CRITICAL: Puppet has not run in the last 10 hours [03:46:01] PROBLEM - Host search32 is DOWN: PING CRITICAL - Packet loss = 100% [03:48:16] RECOVERY - Host search32 is UP: PING OK - Packet loss = 0%, RTA = 0.55 ms [04:00:07] PROBLEM - Host search32 is DOWN: PING CRITICAL - Packet loss = 100% [04:02:12] RECOVERY - Host search32 is UP: PING OK - Packet loss = 0%, RTA = 0.27 ms [04:29:12] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [04:34:45] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.35 ms [04:53:27] PROBLEM - Puppet freshness on neon is CRITICAL: Puppet has not run in the last 10 hours [05:18:30] PROBLEM - Puppet freshness on ocg3 is CRITICAL: Puppet has not run in the last 10 hours [05:20:54] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [05:26:27] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.47 ms [06:03:12] PROBLEM - Puppet freshness on owa1 is CRITICAL: Puppet has not run in the last 10 hours [06:13:15] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [06:18:48] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.41 ms [06:28:02] moring [06:32:36] PROBLEM - Host search32 is DOWN: PING CRITICAL - Packet loss = 100% [07:07:20] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [07:13:02] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.35 ms [07:31:02] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [07:36:44] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.44 ms [07:47:21] bllahh hello [07:51:20] PROBLEM - Puppet freshness on nfs2 is CRITICAL: Puppet has not run in the last 10 hours [07:54:20] PROBLEM - Puppet freshness on nfs1 is CRITICAL: Puppet has not run in the last 10 hours [08:00:02] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [08:05:35] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.51 ms [08:24:02] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [08:29:35] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.38 ms [08:37:01] New patchset: Hashar; "fix syntax errors" [operations/puppet] (production) - https://gerrit.wikimedia.org/r/16011 [08:37:32] apergos: mark: paravoid: mutante: can one of you please merge some syntax errors fixes to puppet ? https://gerrit.wikimedia.org/r/16011 [08:37:36] New review: gerrit2; "Lint check passed." [operations/puppet] (production); V: 1 - https://gerrit.wikimedia.org/r/16011 [08:38:40] Change merged: Faidon; [operations/puppet] (production) - https://gerrit.wikimedia.org/r/16011 [08:38:46] I was looking at the TMH stuff and wondering [08:39:08] thx! [08:39:15] paravoid: which stuff were you looking at ? [08:39:24] timedmediahandler.pp [08:39:29] I'm not very happy about it [08:39:59] the main issue is that it rely on an external repo I guess [08:40:20] that [08:40:24] the lsbdistcodename == lucid [08:40:28] the fact that it's a root class [08:41:48] https://launchpad.net/~j/+archive/timedmediahandler/+packages that Jan packages list [08:41:59] he probably need some specific versions of libs that were not in Lucid [08:42:10] anyway, we will need Precise for TMH [08:42:20] cause the transcoding require a recent ffmpeg package [08:42:28] the one provided with Lucid just don't work [08:43:30] I remember :) [08:43:42] I involved in that, remember? :) [08:44:07] of course :-] [08:44:26] just making sure your brain is still up to date following the 2 weeks of DebianConf brainwashing ! [08:44:28] ;-D [08:44:32] hehehe [08:44:42] I have also replied to your post about puppet tests / doc etc [08:45:07] we have a rake file already https://gerrit.wikimedia.org/r/15568 [08:45:36] looks like generating the doc is all about running something like: puppet doc --mode rdoc --outputdir doc --manifestdir manifests [08:45:36] I just replied to that too [08:45:52] ohh [08:46:58] we could definitely have Jenkins to run puppet lint on the files that got changed [08:47:06] and trigger a doc build every day [08:49:35] paravoid: do you have any knowledge about varnish and our varnish manifest? [08:49:53] I need to set up a cache for upload.beta.wmflabs.org [08:50:18] I know varnish, however I know little about our setup [08:50:35] will dig in our pp files :-] [08:54:23] paravoid: are you in the mood to merge some of my pending changes ? :D [08:54:42] I get a cleanup one https://gerrit.wikimedia.org/r/#/c/15882/ [08:54:56] a if / else hack that was meant for labs, no more needed [08:55:40] hmm maybe it is indeed [08:55:43] that site eqiad [08:55:44] grmbmblblbl [08:56:53] Change abandoned: Hashar; "That is meant to install varnish in eqiad only and squid in other clusters (like pmtpa / knams). So ..." [operations/puppet] (production) - https://gerrit.wikimedia.org/r/15882 [09:02:03] paravoid: do you know how to recover files from the corrupted labs instances ? I would need the /etc/squid/* files from deployment-cache-upload ( i-00000263 ) [09:02:11] if not I will mail Ryan and cross fingers :-] [09:05:25] where are those images? [09:05:38] if he's done it before I'd prefer it if he does it tbh [09:05:45] fine to me [09:05:48] mailing him :-] [09:05:55] if not, I'll be happy to investigate [09:06:09] i am pretty sure he did before [09:06:14] I've heard very little of the specifics of the incident [09:07:23] same to me [09:07:31] so better to let Ryan handle that request. I have mailed him [09:17:40] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [09:23:13] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.41 ms [09:44:01] paravoid: if you are still around, there is also my rakefile enhancement https://gerrit.wikimedia.org/r/#/c/15568/ , let you validate the whole manifests with just 'rake validate' [09:44:11] and add a help message as a default target :) [10:38:20] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [10:43:53] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.38 ms [10:53:32] New review: Hashar; "Stepping out, letting ops review the change." [operations/puppet] (production) C: 0; - https://gerrit.wikimedia.org/r/13484 [11:13:41] PROBLEM - Puppet freshness on virt1002 is CRITICAL: Puppet has not run in the last 10 hours [11:16:41] PROBLEM - Puppet freshness on virt1001 is CRITICAL: Puppet has not run in the last 10 hours [11:22:41] PROBLEM - Puppet freshness on maerlant is CRITICAL: Puppet has not run in the last 10 hours [11:28:50] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [11:34:23] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.35 ms [12:21:46] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [12:21:55] PROBLEM - Puppet freshness on db29 is CRITICAL: Puppet has not run in the last 10 hours [12:23:02] hashar: fwiw, I'm completely lost on the things you need me for :) [12:23:21] (not saying that you did something wrong, on the contrary, I got lost between them) [12:23:38] I guess we are both lost [12:23:49] I have been doing heavy context switching for the last 3 months or so [12:24:00] and I am lost myself [12:26:23] paravoid: I guess we could look at my pending puppet changes ? [12:27:19] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.42 ms [12:29:28] New patchset: Hashar; "(bug 37076) `lint` tool require php5-lint" [operations/puppet] (production) - https://gerrit.wikimedia.org/r/13048 [12:30:04] New review: Hashar; "Added dependency to lint.php script as well." [operations/puppet] (production) C: 0; - https://gerrit.wikimedia.org/r/13048 [12:30:05] New patchset: Hashar; "(bug 37076) `lint` tool require php5-lint" [operations/puppet] (production) - https://gerrit.wikimedia.org/r/13048 [12:30:39] New review: gerrit2; "Lint check passed." [operations/puppet] (production); V: 1 - https://gerrit.wikimedia.org/r/13048 [12:46:49] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [13:08:08] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.37 ms [13:34:50] PROBLEM - Puppet freshness on ms3 is CRITICAL: Puppet has not run in the last 10 hours [13:42:27] !log dist-upgrade and reboot hosts in payments cluster [13:42:35] Logged the message, Master [13:42:37] hiyaaa [13:42:46] anyone there have gerrit project delete powers? [13:43:00] i created my new project repo with an empty first commit [13:43:03] which I did not want to do [13:43:15] and now i'm having trouble pushing my local repository (which already has a few commits) to the remote [13:45:35] Probably need/want Chad or Ryan [13:45:54] hm, ok, is Chad chad on IRC? [13:46:12] He's not here atm but he's ^deamon [13:47:41] ottomata: what is the repo ? maybe I can help [13:48:26] operations/debs/nodejs [13:48:31] https://gerrit.wikimedia.org/r/#/admin/projects/operations/debs/nodejs [13:48:37] <-- hashar btw [13:48:37] you can just delete it, and then I can recreate it properly [13:48:47] hashar! aka gerrit-helper [13:48:49] heheh [13:49:21] @search hashar [13:49:21] Results (Found 1): hashar, [13:49:26] !hashar [13:49:27] [10:15:12] !log WMFLabs seems to have recovered now [13:49:36] :o [13:49:58] ... [13:50:07] I am pretty sure I have whatever bot is responding in my ignore list :-] [13:50:22] heh [13:50:36] ottomata: you could push your own repo using git push -f [13:50:42] ottomata: need a force push right though [13:50:44] -f eh? [13:50:45] hmm [13:50:46] ok [13:51:16] quick q [13:51:19] ottomata: which apparently any people from the analtyics team can do [13:51:21] does this look right? [13:51:22] remote.origin.url=ssh://otto@gerrit.wikimedia.org:29418/operations/debs/nodejs.git [13:51:23] remote.origin.fetch=+refs/heads/*:refs/remotes/origin/* [13:51:23] branch.master.merge=refs/heads/master [13:52:03] please tell me you're not building packages from scratch :) [13:52:21] hehe [13:52:24] nope [13:52:25] not me anyway [13:52:38] john du hart had already done this: [13:52:48] http://svn.mediawiki.org/viewvc/mediawiki/trunk/debs/nodejs/ [13:52:52] oh and I would expect a project under operations/debs to be under ops control [13:53:07] i am just trying to build a nodejs 0.8.2 .deb [13:53:11] yeah, i could do that [13:53:11] ohh [13:53:18] and figured we can change that later [13:53:19] the package in our svn is quiet old [13:53:23] latest and greatest? :) [13:53:26] yeah but it is just the debian [13:53:28] was snapshoted from debian IIRC [13:53:34] isn't there a fresh one in Ubuntu ? [13:53:42] there's an 0.6.x something [13:53:46] afai can tell [13:53:52] (or maybe Ubuntu is going to drop node.js because it release too often *grin* ) [13:53:52] we need at least 0.8.0 [13:53:53] yes [13:54:24] i would put ops perms on this repo, but then I couldn't push straight to it :p, and I'd have to wait long times to get reviews for things that have already been reviewed [13:54:25] and quantal has just 0.6.19 :/ [13:54:32] or i'd just not bother importing this from svn [13:54:34] and just commit to svn [13:54:37] but I do not have svn perms [13:54:39] to commit [13:54:45] forget about svn [13:55:10] ooo, -f might have worked... [13:55:21] yes! [13:55:23] paravoid: don't ubuntu has something like Debian alioth where devs do their packaging work ? [13:55:33] yes! [13:55:34] it works [13:55:38] thanks hashar! [13:55:53] hashar: you mean a thing called "launchpad"? :) [13:55:55] * hashar write down ottomata owe him a beer. [13:56:20] yeah, and there is a guy that has already made these debs in launchpad [13:56:23] i was advised to build our own though... [13:56:35] paravoid: hahh that one indeed [13:56:50] http://ppa.launchpad.net/chris-lea/node.js/ubuntu/pool/main/n/ [13:57:13] paravoid: I though Debian had some kind of git repository for the debian directories [13:57:24] hashar: not exactly [13:57:42] when it comes to choices, Debian picks all of them [13:57:49] where the awesome deb developers would experiment their new packages before sending them to experimental / unstable [13:58:10] so, we have cvs, svn, hg, git, bzr, arch, darcs, monotone [13:58:29] most packages used to use svn though, with git being prevalent nowadays I think [13:58:41] most of them are hosted on git.debian.org, although there's no strict requirement to do so [14:00:01] ahhh http://anonscm.debian.org/gitweb/?p=collab-maint/nodejs.git [14:00:05] that is what I was looking for [14:00:11] from the PTS http://packages.qa.debian.org/n/nodejs.html [14:00:17] collab-maint is a special alioth project where all DDs have commit access [14:00:51] so that is for packages which are simple enough or have no clear maintainer ? [14:01:13] they do have clear maintainers but it's being used either on very simple packages where it doesn't make sense to create a dedicated team [14:01:24] (I love Debian but I barely understand their organization, we are lucky to have you around!) [14:01:29] or on packages that the maintainers feel particularly open about their maintainance [14:01:47] it's an easy way to collaborate, rather than go through the whole process of building a new team [14:02:08] traditionally Debian used to have sole maintainers or e.g. two-three maintainers per package [14:02:20] we're migrating to team maintainance or even collaborative maintainance gradually [14:02:33] in Ubuntu for example, everyone can commit everywhere [14:02:44] both paradigms have their pros and cons [14:03:08] so maybe we could have the node.js packaging done upstream with the Debian folks ? [14:03:15] on the collar-maint/nodejs.git repo? [14:04:47] of course you could [14:04:58] k was wondering. thx [14:04:58] ;) [14:05:12] fresh air timem [14:05:16] we have some sun outside :-] [14:05:20] (which is rare) [14:05:53] PROBLEM - Host virt1002 is DOWN: PING CRITICAL - Packet loss = 100% [14:07:59] RECOVERY - SSH on virt1002 is OK: SSH OK - OpenSSH_5.9p1 Debian-5ubuntu1 (protocol 2.0) [14:08:08] RECOVERY - Host virt1002 is UP: PING OK - Packet loss = 0%, RTA = 35.70 ms [14:08:16] hate gerrit hate [14:08:23] (sorry, I just want to rant) [14:08:24] Change merged: Reedy; [operations/mediawiki-config] (master) - https://gerrit.wikimedia.org/r/15900 [14:09:00] Change merged: Reedy; [operations/mediawiki-config] (master) - https://gerrit.wikimedia.org/r/15807 [14:09:18] please add your hate to the gerrit replacement page [14:09:26] then you can feel like your hate is productive too [14:09:53] heh [14:10:25] I've added two or three entries there [14:10:52] FF14... [14:12:09] what about it? [14:12:19] New patchset: Andrew Bogott; "Temporary change" [operations/puppet] (production) - https://gerrit.wikimedia.org/r/16025 [14:12:28] It's amsuing how quickly they're bumping the major versions now [14:12:55] Change merged: Andrew Bogott; [operations/puppet] (production) - https://gerrit.wikimedia.org/r/16025 [14:14:17] six-weeks isn't it? [14:14:28] six-week release cycle that is [14:14:33] they do have the ESR releases though [14:14:40] Sounds about right [14:15:48] the ESRs are yearly or something [14:28:26] New patchset: Andrew Bogott; "Revert "Temporary change"" [operations/puppet] (production) - https://gerrit.wikimedia.org/r/16028 [14:29:02] Change merged: Andrew Bogott; [operations/puppet] (production) - https://gerrit.wikimedia.org/r/16028 [14:32:10] PROBLEM - Host virt1003 is DOWN: PING CRITICAL - Packet loss = 100% [14:37:43] RECOVERY - Host virt1003 is UP: PING OK - Packet loss = 0%, RTA = 35.66 ms [14:42:04] PROBLEM - SSH on virt1003 is CRITICAL: Connection refused [14:46:51] hashar: around? [14:47:00] paravoid: yup [14:48:08] I'm looking at your review requests [14:48:22] have a few mins to help me understand them? [14:49:05] of course, go ahead! ;-] [14:49:19] so, rakefile; is it used by anything else but users? [14:49:37] why do we even have that? how is it different than "puppet-lint foo.pp" or "puppet parser validate foo.pp"? [14:50:37] lot of newbies don't know about those commands [14:50:53] my point is to provide an easy to use script to people [14:51:02] okay [14:51:05] I don't mind [14:51:05] + puppet-lint comes by default with rules that we do not respect at all [14:51:19] so I wanted to make an easy way to ignore some upstream rules [14:51:22] well, I hope we will with modules [14:51:31] tie the cleanups with the split up [14:51:41] have ops agreed on a set of rules? [14:51:49] Change merged: Faidon; [operations/puppet] (production) - https://gerrit.wikimedia.org/r/15568 [14:51:51] I know MediaWiki people use hard tabs as [14:52:05] whereas I think puppet would like to use 2 spaces [14:52:14] not a big deal though [14:52:24] Change merged: Faidon; [operations/puppet] (production) - https://gerrit.wikimedia.org/r/13339 [14:52:30] yeah!!! [14:53:10] anyway, I wanted to clean out some lintng errors then someone (maybe mark?) told me not to cause it clutter the blame history [14:53:17] which is a valid point [14:53:22] re: lint, that needs manual steps, right? [14:53:43] I am not sure I understand your question? [14:53:57] do you mean it need manual steps to fix the errors reported by lint? answer is yes [14:54:10] though there might some tool around to automatically solve them [14:54:12] which I doubt [14:54:13] sorry, I context switched at a very bad time [14:54:24] I meant php5-lint :) [14:54:25] ssh paravoid vmstat 1 [14:54:31] that needs pecl uninstall, right? [14:54:55] not sure [14:54:58] PROBLEM - Puppet freshness on neon is CRITICAL: Puppet has not run in the last 10 hours [14:55:05] hey boys, q about packaging again [14:55:09] http://wikitech.wikimedia.org/view/Apt [14:55:11] says: [14:55:19] "It's important that we always specify the right distribution (hardy-wikimedia or lucid-wikimedia) for which the package is built in the package's Changelog (debian/changelog)." [14:55:29] !gerrit 13048 [14:55:29] https://gerrit.wikimedia.org/ [14:55:34] er [14:55:36] does that mean that the changelog rev should say 'lucid-wikimedia' in the version? [14:55:46] I disagree that it's important [14:55:49] paravoid: the lint script just load a php script that depends on parsekit_compile_file() provided by the php5-parsekit package you installed. Not sure pecl is related there [14:56:03] iv [14:56:07] i've got this in my control file right now [14:56:07] paravoid: anyway PHP uses a building PECL lib iirc [14:56:07] nodejs (0.8.2-1wm1) lucid-wikimedia; urgency=low [14:56:14] then over at http://wikitech.wikimedia.org/view/Pbuilder [14:56:18] ottomata: that's what it means [14:56:20] whne I create a distribution [14:56:26] should I say [14:56:30] pbuilder create --distribution lucid-wikimedia [14:56:31] ? [14:56:52] hashar: you mention on your commit message something about pecl uninstall etc. [14:57:04] paravoid: hooo yeah [14:57:21] which servers is that? fenari/bast1001? [14:57:23] paravoid: so since we did not have php5-parsekit yet, we had to install that extension using pecl the php extension installer. [14:57:32] New patchset: Faidon; "(bug 37076) `lint` tool require php5-lint" [operations/puppet] (production) - https://gerrit.wikimedia.org/r/13048 [14:57:43] paravoid: whenever debian does not provide a php5-foobar package, we have to pecl install foobar [14:58:06] yeah I got that [14:58:07] paravoid: once 13048 it merged, it can be applied to fenari [14:58:09] New review: gerrit2; "Lint check passed." [operations/puppet] (production); V: 1 - https://gerrit.wikimedia.org/r/13048 [14:58:26] paravoid: the cleanup task would be to remove the old copy that was installed by pecl [14:58:34] Change merged: Faidon; [operations/puppet] (production) - https://gerrit.wikimedia.org/r/13048 [14:58:44] paravoid: I have added the "how to" in commit message as a reminder to whoever was going to merge it [14:59:20] fenari$ pecl list [14:59:21] parsekit 1.3.0 stable [14:59:22] ; [14:59:22] ) [14:59:51] great [15:00:13] PROBLEM - NTP on virt1003 is CRITICAL: NTP CRITICAL: No response from NTP server [15:00:20] so that one could get removed [15:00:24] whenever puppet has run on fenari [15:00:38] not sure if this is relevant, but someone has written a providers for puppet's package resource for pear and pecl [15:00:38] http://www.mit.edu/~marthag/puppet/ [15:00:45] i've used them before and they worked great [15:00:46] please don't [15:01:06] package { "foobar": provider => "pecl" } [15:01:14] :-D [15:01:24] ops don't want to have puppet to install from third parties repository [15:01:29] aye cool [15:01:29] I think that installing whatever packaging each language decided to create is a very bad idea [15:01:36] ok, fine with me! [15:01:52] someone could sneak a faulty package in upstream repo and kill us :-) [15:01:56] I guess that is the main point [15:02:08] I've worked with systems that had gems, pypi, pecl and whatnot [15:02:23] they sneak bugs instead, bastards! [15:02:38] every language author thinks they're better than distributions [15:02:42] and they can do a better job [15:03:05] and that it's /their/ job of providing tools for the sysadmins [15:03:06] ottomata: if you ever need a specific php5 extension, it is best to use the debian package. If it is not there we can get ops to build the .deb package for us (Debian has a script that generate a nice package out of pecl. Something like : dh-build-from-pecl foobar) [15:03:23] php is even more fucked up, they decided to create /two/ of them! [15:03:31] ah ok cool [15:03:33] composer!!! ;-D [15:03:45] good to konw [15:03:52] for php http://getcomposer.org/ [15:04:12] paravoid, do you remember what it was like when you were learning to create .debs for the first time [15:04:13] ? [15:04:14] AGH! [15:04:18] SO MANY DIFFERENT WAYS [15:04:19] hehe [15:04:24] q [15:04:30] if I have a debian/ dir that works [15:04:31] I only managed to build two so far :-/ [15:04:32] can I just run [15:04:36] dpkg-buildpackage [15:04:37] ? [15:04:38] and both time had to find a doc [15:04:44] anyway, now I ask paravoid :-] [15:04:45] or do I have to use pbuilder? [15:04:52] do I have to have a chroot? [15:05:08] ottomata: if you're a newbie, I wouldn't recommend to you to build packages for nodejs [15:05:10] echo 'can you please build a package for foo? thanks -- antoine' | mail faidon@wikimedia.org [15:05:17] well i don't have to create the debian/ [15:05:20] john du hart already has [15:05:26] and I build it locally [15:05:29] seemd to work great [15:05:42] all I did was change changelog, and change a dependency in debian/control [15:05:50] most of the work is doing by nodejs's build scripts [15:06:07] http://svn.mediawiki.org/viewvc/mediawiki/trunk/debs/nodejs/debian/rules?revision=101926&view=markup [15:06:17] half of upstream build systems are completely fucked up :) [15:06:28]