[00:29:08] (03PS3) 10BryanDavis: Revert "DNS cleanup" [operations/dns] - 10https://gerrit.wikimedia.org/r/120039 [01:38:54] PROBLEM - Puppet freshness on labstore2 is CRITICAL: Last successful Puppet run was Fri 21 Mar 2014 01:17:26 AM UTC [02:17:32] !log LocalisationUpdate completed (1.23wmf19) at 2014-03-24 02:17:32+00:00 [02:17:44] Logged the message, Master [02:41:16] !log LocalisationUpdate ResourceLoader cache refresh completed at Mon Mar 24 02:41:13 UTC 2014 (duration 41m 12s) [02:41:21] Logged the message, Master [03:02:02] (03PS1) 10Springle: repool db1037 in s5 [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120484 [03:02:39] (03CR) 10Springle: [C: 032] repool db1037 in s5 [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120484 (owner: 10Springle) [03:02:47] (03Merged) 10jenkins-bot: repool db1037 in s5 [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120484 (owner: 10Springle) [03:03:51] !log springle synchronized wmf-config/db-eqiad.php 's5 repool db1037 warm up' [03:03:56] Logged the message, Master [04:39:54] PROBLEM - Puppet freshness on labstore2 is CRITICAL: Last successful Puppet run was Fri 21 Mar 2014 01:17:26 AM UTC [08:29:17] hello! [08:29:35] Openstack has a puppet dashboard: http://puppetdb.openstack.org/ :-D [08:33:18] hi hashar [08:33:49] lo [08:51:04] RECOVERY - HTTP 5xx req/min on tungsten is OK: OK: reqstats.5xx [warn=250.000 [08:53:37] (03PS2) 10Hashar: Fix error: timidity service can not be stopped [operations/puppet] - 10https://gerrit.wikimedia.org/r/118709 [08:53:52] (03CR) 10Hashar: "Mentioned in commit summary message:" [operations/puppet] - 10https://gerrit.wikimedia.org/r/118709 (owner: 10Hashar) [08:54:29] (03CR) 10Hashar: "> looks sane, i would wait with those changes until pmtpa is out." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120013 (owner: 10Hashar) [08:55:57] (03CR) 10Hashar: "> did you see the request at the top of this file? Please abandon." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120005 (owner: 10Hashar) [08:57:09] (03CR) 10Matanya: "fair enough, tim and hashar, i accept your view." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120005 (owner: 10Hashar) [09:58:41] (03PS1) 10ArielGlenn: snapshots: move centralauthdump cron out of role into module [operations/puppet] - 10https://gerrit.wikimedia.org/r/120496 [10:00:39] (03CR) 10ArielGlenn: [C: 032] snapshots: move centralauthdump cron out of role into module [operations/puppet] - 10https://gerrit.wikimedia.org/r/120496 (owner: 10ArielGlenn) [10:07:19] akosiaris: ping [10:07:39] (03PS1) 10Hashar: contint: install puppet-lint from rubygems on labs [operations/puppet] - 10https://gerrit.wikimedia.org/r/120498 [10:08:28] (03CR) 10jenkins-bot: [V: 04-1] contint: install puppet-lint from rubygems on labs [operations/puppet] - 10https://gerrit.wikimedia.org/r/120498 (owner: 10Hashar) [10:09:33] (03PS2) 10Hashar: contint: install puppet-lint from rubygems on labs [operations/puppet] - 10https://gerrit.wikimedia.org/r/120498 [10:11:17] (03CR) 10Matanya: "wouldn't it be better to backport a newer version? gem2deb is your friend :)" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120498 (owner: 10Hashar) [10:12:23] (03PS1) 10Hashar: snapshot: lint list-last-n-good-dumps.py [operations/puppet] - 10https://gerrit.wikimedia.org/r/120499 [10:13:38] (03CR) 10jenkins-bot: [V: 04-1] snapshot: lint list-last-n-good-dumps.py [operations/puppet] - 10https://gerrit.wikimedia.org/r/120499 (owner: 10Hashar) [10:13:49] bahh [10:15:17] (03PS2) 10Hashar: snapshot: lint list-last-n-good-dumps.py [operations/puppet] - 10https://gerrit.wikimedia.org/r/120499 [10:19:02] (03Abandoned) 10Hashar: Move logs to /var/log/mediawiki [operations/puppet] - 10https://gerrit.wikimedia.org/r/83574 (owner: 10Reedy) [10:20:42] (03Abandoned) 10Hashar: Add csteipp & aaron to admins::jenkins group [operations/puppet] - 10https://gerrit.wikimedia.org/r/119209 (owner: 10Ori.livneh) [10:37:26] (03PS2) 10Hashar: openstack: generic_upstart now use boolean values [operations/puppet] - 10https://gerrit.wikimedia.org/r/118716 [10:37:29] (03PS2) 10Hashar: lvs: generic_upstart now use boolean values [operations/puppet] - 10https://gerrit.wikimedia.org/r/118717 [10:37:32] (03PS2) 10Hashar: twemproxy: generic_upstart now use boolean values [operations/puppet] - 10https://gerrit.wikimedia.org/r/118718 [10:41:54] PROBLEM - Puppet freshness on labstore2 is CRITICAL: Last successful Puppet run was Fri 21 Mar 2014 01:17:26 AM UTC [10:44:37] (03PS3) 10Hashar: twemproxy: generic::upstart_job() now uses boolean values [operations/puppet] - 10https://gerrit.wikimedia.org/r/118718 [10:44:47] (03PS3) 10Hashar: openstack: generic::upstart_job() now uses boolean values [operations/puppet] - 10https://gerrit.wikimedia.org/r/118716 [10:45:04] (03PS3) 10Hashar: lvs: generic::upstart_job() now uses boolean values [operations/puppet] - 10https://gerrit.wikimedia.org/r/118717 [10:47:55] (03PS10) 10Hashar: sanity test for refreshWikiversionsCDB [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/105698 [10:49:46] (03CR) 10Matanya: lvs: generic::upstart_job() now uses boolean values (031 comment) [operations/puppet] - 10https://gerrit.wikimedia.org/r/118717 (owner: 10Hashar) [10:58:48] (03CR) 10Matanya: [C: 031] openstack: generic::upstart_job() now uses boolean values (032 comments) [operations/puppet] - 10https://gerrit.wikimedia.org/r/118716 (owner: 10Hashar) [10:59:51] (03CR) 10Matanya: twemproxy: generic::upstart_job() now uses boolean values (031 comment) [operations/puppet] - 10https://gerrit.wikimedia.org/r/118718 (owner: 10Hashar) [12:21:43] (03PS1) 10Matanya: puppetmaster: qualify var [operations/puppet] - 10https://gerrit.wikimedia.org/r/120512 [12:30:04] PROBLEM - HTTP 5xx req/min on tungsten is CRITICAL: CRITICAL: reqstats.5xx [crit=500.000000 [12:44:29] (03CR) 10Faidon Liambotis: [C: 032] Revert "DNS cleanup" [operations/dns] - 10https://gerrit.wikimedia.org/r/120039 (owner: 10BryanDavis) [12:46:14] the reqerror spike is one user with UA PECL::HTTP/1.7.4 (PHP/5.3.22) [12:46:21] with some API gets [12:57:04] wtf is wrong with ulsfo [12:57:31] icinga was all red for 4xxx for a moment there [12:58:44] connectivity flapped? [12:59:12] if it did, it was short enough to not be noticed by OSPF/BGP [13:14:22] (03PS1) 10Matanya: mha: fix var scope search [operations/puppet] - 10https://gerrit.wikimedia.org/r/120518 [13:18:36] (03PS2) 10Matanya: mha: fix var scope search [operations/puppet] - 10https://gerrit.wikimedia.org/r/120518 [13:29:19] * paravoid grumbles at the icinga labs warnings [13:35:35] (03PS1) 10Faidon Liambotis: Fold exim::rt into role::rt [operations/puppet] - 10https://gerrit.wikimedia.org/r/120520 [13:35:37] (03PS1) 10Faidon Liambotis: spamassassin: move into a module [operations/puppet] - 10https://gerrit.wikimedia.org/r/120521 [13:35:39] (03PS1) 10Faidon Liambotis: spamassassin: cleanup module [operations/puppet] - 10https://gerrit.wikimedia.org/r/120522 [13:35:41] (03PS1) 10Faidon Liambotis: clamav: move into a module [operations/puppet] - 10https://gerrit.wikimedia.org/r/120523 [13:42:54] PROBLEM - Puppet freshness on labstore2 is CRITICAL: Last successful Puppet run was Fri 21 Mar 2014 01:17:26 AM UTC [13:43:17] (03PS2) 10Faidon Liambotis: clamav: move into a module [operations/puppet] - 10https://gerrit.wikimedia.org/r/120523 [13:43:19] (03PS2) 10Faidon Liambotis: spamassassin: cleanup module [operations/puppet] - 10https://gerrit.wikimedia.org/r/120522 [13:48:21] whoah, spamassassin improvements [13:48:28] not really [13:48:48] (preparation for) improvements [13:48:55] TLC [13:48:57] whatever [13:49:04] do you have something specific in mind? [13:49:07] that you'd like to see fixed I mean [13:49:11] (03CR) 10Faidon Liambotis: [C: 032] Fold exim::rt into role::rt [operations/puppet] - 10https://gerrit.wikimedia.org/r/120520 (owner: 10Faidon Liambotis) [13:49:31] (03CR) 10Faidon Liambotis: [C: 032] "Catalog-diffed." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120521 (owner: 10Faidon Liambotis) [13:49:39] (03CR) 10Faidon Liambotis: [C: 032] "Catalog-diffed." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120522 (owner: 10Faidon Liambotis) [13:49:48] (03CR) 10Faidon Liambotis: [C: 032] "Catalog-diffed." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120523 (owner: 10Faidon Liambotis) [13:54:28] (03PS1) 10Faidon Liambotis: otrs: add missing network::constants include [operations/puppet] - 10https://gerrit.wikimedia.org/r/120526 [13:54:39] (03CR) 10Faidon Liambotis: [C: 032] otrs: add missing network::constants include [operations/puppet] - 10https://gerrit.wikimedia.org/r/120526 (owner: 10Faidon Liambotis) [13:54:49] paravoid: the spamd stuff [13:55:00] yes? [13:55:12] last i asked it was broken [13:55:19] what does that mean? [13:55:37] i remember i spoke with akosiaris about it some time ago [13:55:57] akosiaris: https://gerrit.wikimedia.org/r/#/c/117024/ :) [13:56:16] he said something about the hoem dir and the usage of systemuser [13:56:24] he's on vacation this week [13:56:36] matanya: do you see those in my changeset? :) [13:56:46] see which? [13:56:58] paravoid: akosiaris i? [13:56:59] is? [13:57:04] the broken /var/spamd and systemuser [13:57:11] ottomata: yes, iirc [13:57:14] ah, hmmmm [13:57:18] paravoid: no, where is it ? [13:57:26] * matanya feels stupid [13:57:35] who is less buys than you and wants to review it? the main part I feel like needs review is the rsync daemon [13:57:38] on a public IP [13:57:55] * matanya found it [13:58:18] well, paravoid that :P [13:59:02] thanks for fixing it [13:59:23] (03PS13) 10Ottomata: Adding archiva module and role, applying on titanium [operations/puppet] - 10https://gerrit.wikimedia.org/r/117024 [14:10:33] aude: I see you have gerrit change 119311 on for the SWAT deploy in about 50 minutes. But it's not merged to master yet. [14:13:50] working on it [14:14:01] i suppose if not in yet, then we'll do later [14:14:30] i can submit core submodule patch [14:23:29] ottomata: still on for 1030? [14:24:22] oh yeahhhhhH! [14:24:32] i forgot but now that you rminded me i'm excited! [14:25:09] great...one at a time? an1015 first [14:25:22] yeah one at a time [14:26:24] wait, 1015? [14:26:33] looking at RT ticket [14:26:40] We should move analytics 1018,1019,1020, and 1025 from Row C into Row D. [14:27:00] cmjohnson1: ^ [14:27:03] ok [14:28:37] ok yeah, so there are 7 out of 10 datanodes in row c right now [14:28:52] which means there are 3 in row a [14:28:56] so [14:29:01] we want to balance as best as possible [14:29:08] so we are going to move 3 out of C and into D [14:29:12] that will leave us at [14:29:18] A: 3 [14:29:19] C: 4 [14:29:19] D: 3 [14:29:34] so, 1018,1019,1020 will move to Row D [14:29:42] got it [14:30:00] cool, so let's do 1018! [14:30:05] you ready? i will go in and shut it down [14:30:10] yes, i am ready [14:31:51] !log stopping hadoop services and shutting down analytics1018 [14:31:57] Logged the message, Master [14:32:37] cmjohnson1: it should be shutting down now [14:33:07] (03CR) 10Addshore: [C: 031] Remove language subdomains for wikidata.org [operations/dns] - 10https://gerrit.wikimedia.org/r/119032 (owner: 10Faidon Liambotis) [14:34:24] PROBLEM - Host analytics1018 is DOWN: PING CRITICAL - Packet loss = 100% [14:40:52] (03CR) 10Ottomata: [C: 032 V: 032] Initial 2.0.0-1 debian release [operations/debs/archiva] (debian) - 10https://gerrit.wikimedia.org/r/115323 (owner: 10Ottomata) [14:46:45] aude: Looking at these Wikibase changes for the SWAT, I'm a little lost because I can't actually find where mediawiki/extensions/Wikibase is actually deployed on the WMF servers. I see extensions/Wikidata/extensions/Wikibase seems to have a static copy of it, but that doesn't help me much. [14:47:27] anomie: Yep it's in the Wikidata extension [14:47:45] We use a copy in there with all of its dependencies [14:47:56] it's in Wikidata [14:48:27] i suppose we can wait until the later slot tonight, since we're still organizing these patches [14:48:37] ottomata: is the Jenkins debian-glue job any helpful ? :-] [14:48:48] hashar: uhhh [14:49:09] i have never looked at it :-$ [14:49:12] ohhh [14:49:26] yep, it's probably better if we wait [14:49:44] * aude might need nap, but ok [14:49:59] I am asking because some init script lintian error got fixed with PS 12 https://gerrit.wikimedia.org/r/#/c/115323/11..12/debian/archiva.init [14:50:35] and the job did mention them at https://integration.wikimedia.org/ci/job/operations-debs-archiva-debian-glue/13/testReport/ [14:50:46] ahh, i think whatever that was was also mentioned in review [14:50:51] oh and I see the links in the comments now [14:50:59] hmm, cool, I will click on those more often [14:51:23] hey anomie and manybubbles, looks like we've got nothing to deploy today [14:51:24] the whole idea is to report lintian errors automatically [14:51:37] aye, i mean, i see them when I build the package too [14:51:38] so the patch author don't waste reviewers time with things that are easily fixable [14:51:40] so I fix many of them then [14:51:45] but some I don't bother fixing [14:51:47] :-° [14:51:55] now the job does it for ya automatically :] [14:51:56] MaxSem: sounds good to me [14:52:01] I'll schedule something for tomorrow [14:52:13] we might find something for tomorrow [14:52:13] ottomata: great to see you got all lintian errors fixed ! [14:53:05] * anomie is glad for public logs in light of his connection issues this morning [14:53:45] * matanya was pinged by lint* :P [14:54:09] MaxSem: We almost had some Wikidata patches, but it turned out they weren't quite ready [14:54:14] (03PS1) 10Cmjohnson: Updating dns entry for analytics1018 [operations/dns] - 10https://gerrit.wikimedia.org/r/120533 [14:56:30] (03CR) 10Ottomata: [C: 032] Updating dns entry for analytics1018 [operations/dns] - 10https://gerrit.wikimedia.org/r/120533 (owner: 10Cmjohnson) [14:57:19] (03CR) 10Cmjohnson: [C: 032] Updating dns entry for analytics1018 [operations/dns] - 10https://gerrit.wikimedia.org/r/120533 (owner: 10Cmjohnson) [14:59:11] paravoid: I moved to analytics1018 to a different vlan...had to change ip. Can you double check to make sure we don't have any network issues (will also be moving an1019 and 1020) [15:00:14] (03CR) 10Hashar: [C: 031 V: 032] "deployed on beta cluster via local puppet master" [operations/puppet] - 10https://gerrit.wikimedia.org/r/118709 (owner: 10Hashar) [15:00:27] cmjohnson1: if we boot it back on, i'll log in and check some networking stuff [15:01:06] ottomata: powering on...try and login via mgmt [15:01:45] k... [15:01:56] in console, i see it booting [15:02:23] cool...so are you going to change interface settings? [15:02:50] ja will do [15:04:10] okay..ping me when you wanna move 1019 [15:07:22] lemme know if you need anything [15:11:06] (03PS1) 10Hoo man: Run rebuildEntityPerPage.php on Wikidata (once per month) [operations/puppet] - 10https://gerrit.wikimedia.org/r/120535 [15:15:36] cmjohnson1: is 10.64.53.1 the correct gateway? [15:16:00] yes [15:17:22] hm [15:17:49] something isn't working [15:17:56] not yet sure if i've done something wrong [15:18:01] but I can't ping the gateway even [15:20:39] haha, oof, why does this always happen cmjohnson1? :p [15:21:06] did you update /etc/hosts ? [15:21:24] (03Abandoned) 10coren: toollabs: insert sql tool to execnodes [operations/puppet] - 10https://gerrit.wikimedia.org/r/66266 (owner: 10Petrb) [15:21:27] yes [15:22:29] ottomata: is this what you have http://p.defau.lt/?aHr_Tbb_cx_xjZxDT_bt3w [15:22:40] (03PS1) 10coren: Tool Labs: new-style cron via submit host [operations/puppet] - 10https://gerrit.wikimedia.org/r/120538 [15:23:01] yup [15:23:05] cmjohnson1 ^ [15:23:42] http://p.defau.lt/?U_f7niay_Jq_daxPAlPkDw [15:23:44] paravoid: is that correct broadcast? see link [15:23:52] (03PS1) 10ArielGlenn: remove nonexistent param from content retriever invocation (wikiretriever) [operations/dumps] (ariel) - 10https://gerrit.wikimedia.org/r/120539 [15:24:23] (03CR) 10ArielGlenn: [C: 032] remove nonexistent param from content retriever invocation (wikiretriever) [operations/dumps] (ariel) - 10https://gerrit.wikimedia.org/r/120539 (owner: 10ArielGlenn) [15:24:35] (03CR) 10coren: [C: 032] Tool Labs: new-style cron via submit host [operations/puppet] - 10https://gerrit.wikimedia.org/r/120538 (owner: 10coren) [15:24:36] i don't htink broadcast would matter in this case, but even so it should be the correct broadcast with that netmask [15:24:50] pretty sure that is all correct [15:24:53] Coren: Y u so fast? :P [15:25:00] toollabs::execnode also has the sql command [15:25:01] ottomata: the vlan didn't set correctly [15:25:08] oh? [15:25:11] and should have been updated [15:25:20] hoo: Who put that /there/? [15:25:37] finds git blame. [15:25:40] Maybe it was me :P [15:25:43] It was me [15:25:49] Hah-ha! [15:25:50] :-) [15:26:05] Wasn't aware of exec_environ than I put that up [15:26:10] oversight :/ [15:26:35] It was mostly an accident that I merged the sql bit, I forgot to split the diffs and when I did the git commit I didn't want to bother inmixing them up. :-) [15:26:42] Meh, trivial fix. :-0 [15:27:19] Coren: trivial merge? https://gerrit.wikimedia.org/r/#/c/120243/ [15:27:33] Coren: and https://gerrit.wikimedia.org/r/#/c/120348/ [15:27:33] ottomata: something is not right..it won't save to that vlan..maybe mark can offers some advice [15:27:59] (03CR) 10coren: [C: 032] "Trivial fix is trivial." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120243 (owner: 10Tim Landscheidt) [15:29:12] it won't save? [15:29:15] whatcha mean? [15:29:25] (03CR) 10coren: [C: 032] "Less trivial, but also quite reasonable." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120348 (owner: 10Tim Landscheidt) [15:29:46] Coren: ty [15:30:15] what's up? [15:30:36] mark, cmjohnson1 and I are moving nodes to the new analytics vlan in row d [15:30:41] mark: the interface for analytics1018 won't save to analytics vlan [15:30:48] it doesn't give me any errors on commit [15:30:57] yuvipanda: You may want to look at the apache config on tools-webproxy -- there are a few tricks in place still. When there isn't a registered webservice, I serve a specific errordocument and when the tools.admin webservice isn't up, I serve a static version. (And / is handled by the tools.admin webservice) [15:31:15] Coren: ah, right. I'll pick that up. [15:31:17] cmjohnson1: then why do you think so? [15:31:23] yuvipanda: You /could/ simply fall back to the apache webproxy, actually, but I'd rather get rid of it entirely. [15:31:37] Coren: yeah, I want to replace that thing. [15:32:28] (03PS1) 10coren: Tool Labs: avoid double definition of /usr/bin/sql [operations/puppet] - 10https://gerrit.wikimedia.org/r/120540 [15:32:54] ah :) [15:33:18] mark, its possible i've done something wrong when changing network/interfaces and etc/hosts on the box, but i can't ping the gateway, and route takes longer than usual to print routes (not sure what that means) [15:34:08] you haven't removed those hosts from the other vlan [15:34:11] (03CR) 10BryanDavis: "Nice! I had to comment this bit of code out in my local hacks to test scap under Vagrant. I did not try to do nearly as much research as A" [operations/puppet] - 10https://gerrit.wikimedia.org/r/118709 (owner: 10Hashar) [15:34:18] haha..just about to say that [15:34:30] took a minute [15:35:19] [edit interfaces interface-range vlan-public1-d-eqiad] [15:35:19] + member "ge-2/0/[3-47]"; [15:35:19] - member ge-2/0/*; [15:35:29] (03CR) 10coren: [C: 032] Tool Labs: avoid double definition of /usr/bin/sql [operations/puppet] - 10https://gerrit.wikimedia.org/r/120540 (owner: 10coren) [15:36:37] mark: thx [15:37:34] (03PS3) 10Hashar: beta: sent HTCP purges to eqiad varnishes [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/116788 [15:38:50] (03CR) 10Hashar: "I have no idea how timidity-daemon ends up being installed on the production application servers though :-(" [operations/puppet] - 10https://gerrit.wikimedia.org/r/118709 (owner: 10Hashar) [15:40:04] PROBLEM - HTTP 5xx req/min on tungsten is CRITICAL: CRITICAL: reqstats.5xx [crit=500.000000 [15:40:59] (03PS1) 10ArielGlenn: remove searchlogs rsync conf from datasets manifests, no longer used [operations/puppet] - 10https://gerrit.wikimedia.org/r/120543 [15:42:03] mark: are you going to add to analytics1-d-eqiad ? [15:42:13] ah I can [15:42:45] but it was already added? [15:43:06] it is in network.pp [15:43:08] not sure if that is what you mean [15:43:11] (03PS2) 10ArielGlenn: remove searchlogs rsync conf from datasets manifests, no longer used [operations/puppet] - 10https://gerrit.wikimedia.org/r/120543 [15:43:43] oh..still showing under default [15:44:01] oh [15:44:05] the interface range is just not setup [15:44:36] yeah, I can get that...there is a msg that you were configuring still so I didn't wanna mess with it [15:44:47] Users currently editing the configuration: [15:45:09] oh i'm already done [15:45:11] sorry [15:45:25] ottomata: did you see my response re: stefan? [15:45:31] (03CR) 10ArielGlenn: [C: 032] remove searchlogs rsync conf from datasets manifests, no longer used [operations/puppet] - 10https://gerrit.wikimedia.org/r/120543 (owner: 10ArielGlenn) [15:46:17] yes I asked him to respond [15:46:29] oh he just did [15:47:14] ottomata: i can ping 10.64.53.10 now [15:47:22] (03PS1) 10coren: Tool Labs: add role for the new submit host [operations/puppet] - 10https://gerrit.wikimedia.org/r/120546 [15:47:52] great me too! [15:49:22] (03CR) 10coren: [C: 032] "Simple role addition." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120546 (owner: 10coren) [15:50:03] great, looks good cmjohnson1 [15:50:30] cmjohnson1: i'm going to shut down an19, you ready? and then do this key revocation for stefan [15:50:32] while you move [15:50:58] give me a few mins...set it to shut down in 10 [15:51:21] ok [15:51:22] * cmjohnson1 needs coffee still [15:51:26] ok no probs [15:51:29] lemm eknow when [15:53:00] (03PS1) 10Ottomata: Removing Stefan Petrea's old ssh keys, installing new [operations/puppet] - 10https://gerrit.wikimedia.org/r/120548 [15:53:28] (03PS2) 10Ottomata: Removing Stefan Petrea's old ssh keys, installing new [operations/puppet] - 10https://gerrit.wikimedia.org/r/120548 [15:53:42] (03CR) 10Ottomata: [C: 032 V: 032] Removing Stefan Petrea's old ssh keys, installing new [operations/puppet] - 10https://gerrit.wikimedia.org/r/120548 (owner: 10Ottomata) [15:53:53] jesus how large is this key [15:54:05] hah [15:54:33] seriously, how large is it? [15:55:16] asking [15:55:19] 4k? 8k? [15:55:30] probably 8k? [15:55:58] (03CR) 10Lydia Pintscher: [C: 031] Remove language subdomains for wikidata.org [operations/dns] - 10https://gerrit.wikimedia.org/r/119032 (owner: 10Faidon Liambotis) [15:56:14] mwalker: hey, around? [15:56:38] yep [15:56:43] hi [15:56:45] what can I do for you this morning? [15:56:55] https://wikimediafoundation.org/wiki/Thank_you has a Special:HideBanners link for en.wikidata.org [15:56:59] we are about to deprecate that domain [15:57:16] (wikidata is just "wikidata.org" & "www.wikidata.org") [15:57:17] haha, paravoid 16384 [15:57:23] ottomata: yeah, no. [15:58:12] oh; good to know [15:58:24] I shall change the page; and try and track down the other places we probably have that [15:58:42] ottomata: 2K should be enough, 4K if he's feeling really paranoid. revoke the 16K. [15:58:57] (or really, ECDSA if he's feeling paranoid) [15:59:21] sigh, ok... [15:59:23] asking him to join here [15:59:23] mwalker: thanks! I did a quick git grep but couldn't find it anywhere, so it probably was more complicated than that and didn't have the time [15:59:41] mwalker: could you comment on https://gerrit.wikimedia.org/r/119032 when you are done? [15:59:42] joined [15:59:49] hi paravoid , ottomata [15:59:53] hi average [16:00:34] mwalker: would you like me to file a bug perhaps? sorry for being an interrupt like that :) [16:00:55] average, paravoid says your key is too long! [16:01:09] is 8192 bits ok ? [16:01:12] no [16:01:16] 4096 ? [16:01:21] 17:58 < paravoid> ottomata: 2K should be enough, 4K if he's feeling really paranoid. revoke the 16K. [16:01:24] 17:58 < paravoid> (or really, ECDSA if he's feeling paranoid) [16:01:46] note that the more bits you add, the longer it will take for you to login [16:02:07] ok, so I guess you'd agree with 4096 [16:02:14] I'll go for that, if that's ok [16:02:18] most of the roots are 2K too, so I don't see the security benefit of you having 4K tbh [16:02:41] but sure, go for 4K if you want [16:02:43] paranoid much? :) [16:02:48] very(lately) [16:03:11] RECOVERY - Host analytics1018 is UP: PING OK - Packet loss = 0%, RTA = 0.54 ms [16:03:11] PROBLEM - NTP on analytics1018 is CRITICAL: NTP CRITICAL: Offset unknown [16:03:36] ok average put a new key on office and i'll put it in place [16:04:06] (03CR) 10Faidon Liambotis: "Tobias, it's the Special:HideBanners links. I've informed fundraising and they are on it. I'll merge as soon as they're done, hopefully ve" [operations/dns] - 10https://gerrit.wikimedia.org/r/119032 (owner: 10Faidon Liambotis) [16:04:41] ottomata: don't forget to revoke the 16K too now, as it's already being deployed across the infra (unless you didn't get the chance to puppet-merge?) [16:06:20] (03CR) 10Greg Grossmeier: [C: 031] "Also:" [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120170 (owner: 10Spage) [16:06:52] paravoid , paravoid regenerated, uploaded to office [16:06:53] ottomata whenever [16:07:03] yeah i'll set it as absent [16:08:08] ottomata: ^^ [16:08:50] cmjohnson1: ah, i've gotta do this key thing real quick now, and I'm super duper hungry [16:09:01] sorry we are out of sync, could we continue after ops meeting? [16:09:05] or maybe before? [16:09:12] ottomata: sure [16:09:40] average: hm, I htink you need to name your key something different :/ [16:09:40] greg-g: I was thinking that we should special-case loginwiki and votewiki to get nothing by default, possibly. [16:09:56] or, hm, no, do I? [16:10:03] ottomata: what should I name it ? [16:10:11] paravoid: I can just replace the key, right? puppet will replace it (effectively revoking the old one) [16:10:13] right? [16:10:24] no [16:10:25] it won't [16:10:40] it doesn't replace, it just adds [16:10:50] really? even if it is named the same? [16:10:54] I think so [16:10:56] its a define [16:10:59] the name is unique [16:11:07] doesn't matter [16:11:49] James_F: /me nods [16:12:37] ah its not a define, its a resouce [16:12:37] yeah [16:12:38] paravoid [16:12:39] name [16:12:39] (Namevar: If omitted, this attribute’s value defaults to the resource’s title.) [16:12:39] The SSH key comment. This attribute is currently used as a system-wide primary key and therefore has to be unique. [16:12:51] can we figure this out later please? [16:12:56] just ensure => absent it for now [16:12:58] it won't hurt [16:13:00] we can remove it later [16:13:04] ok, average [16:13:07] you need to rename your key then [16:13:12] doesn't matter [16:13:14] it doesn't matter [16:13:22] puppet will error out [16:13:24] the name of the key is arbitrary, you can set it to whatever you want [16:13:25] if they have the same name [16:13:33] it's not encoded in the key itself [16:13:36] the resource is titled by the name [16:13:41] oh no? [16:13:45] the title doesn't matter? [16:13:47] ok [16:13:49] it's just a comment [16:14:19] it doesn't have to be the same on client & server, it's not encoded in the key, it's not part of the exchange at all [16:16:36] just if you were looking at authorized_keys on the server it helps to know which is which :) [16:17:10] (03PS1) 10Ottomata: Revoking Stefan's most recent key and adding a shorter one [operations/puppet] - 10https://gerrit.wikimedia.org/r/120550 [16:17:20] (03PS2) 10Ottomata: Revoking Stefan's most recent key and adding a shorter one [operations/puppet] - 10https://gerrit.wikimedia.org/r/120550 [16:17:22] (03CR) 10jenkins-bot: [V: 04-1] Revoking Stefan's most recent key and adding a shorter one [operations/puppet] - 10https://gerrit.wikimedia.org/r/120550 (owner: 10Ottomata) [16:17:47] (03CR) 10Ottomata: [C: 032 V: 032] Revoking Stefan's most recent key and adding a shorter one [operations/puppet] - 10https://gerrit.wikimedia.org/r/120550 (owner: 10Ottomata) [16:30:53] (03PS1) 10coren: Tool Labs: ensure latest for requested packages [operations/puppet] - 10https://gerrit.wikimedia.org/r/120553 [16:32:24] (03CR) 10coren: [C: 032] "Simple tweak." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120553 (owner: 10coren) [16:38:02] (03PS1) 10coren: Tool Labs: tweaks to submit crontab [operations/puppet] - 10https://gerrit.wikimedia.org/r/120556 [16:41:33] (03CR) 10coren: [C: 032] "That works." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120556 (owner: 10coren) [16:43:07] mutante: Is the 'planet' labs project migrated? Or, if not, do you need any support with that? [16:43:51] PROBLEM - Puppet freshness on labstore2 is CRITICAL: Last successful Puppet run was Fri 21 Mar 2014 01:17:26 AM UTC [16:45:14] (03PS1) 10coren: Tool Labs: minor typo fix [operations/puppet] - 10https://gerrit.wikimedia.org/r/120558 [16:47:23] andrewbogott: i'll likely have it killed before meeting. k? [16:47:34] just checking [16:47:56] mutante: if you just want me to delete it then I can add it to the list right now, you don't need to do anything. [16:48:54] (03CR) 10coren: [C: 032] "Silly brain." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120558 (owner: 10coren) [16:52:06] hell week begins now \o/ ;] [16:54:42] oh hey robh, so, since your on RT duty... [16:56:37] andrewbogott: Categories: [16:56:53] Projects used in production .. not really [16:57:14] ? [16:57:33] https://wikitech.wikimedia.org/wiki/Category:Projects_used_in_production [16:57:49] it's in that category, and I thought that made you treat it differentely, maybe [16:58:17] Nope, it's just on the short list of projects that people have claimed but not finished migrating... [16:58:33] all deadlines have long-since passed so I'm on a nagging rampage [16:59:30] i don't use it anymore but it seems others do, i'll rebort back within 15 min:) [17:01:40] ok [17:14:35] andrewbogott: just shut it all down :P [17:15:15] matanya: I'm actually pretty surprised at how many projects are still active. I was betting on 80% attrition, it's turned out to be more like 50-60%. [17:15:48] i suspect a lot of them are not used anymore [17:15:48] It's really nice that so many people are using labs, and responding quickly and politely to my godzilla-like stomping of pmtpa :) [17:16:09] matanya: no, that's what I mean, I've gotten active responses and participation in the migration of about half the projects. [17:16:38] well, maybe not quite half, I'll count after the dust settles [17:17:15] I'd say the more useful metric is "number of instances in maintained projects" rather than "number of maintained projects" [17:17:44] Ah, true, that's definitely more than 50%. [17:17:47] andrewbogott: deleted instances, removed hostname, released public IP... [17:18:00] re: planet project just has eqiad instance [17:18:17] mutante: Thanks, shall I move it to the 'migration finished' column? [17:18:27] andrewbogott: yes please [17:19:35] if needed i'll request the hostname, but i should never need the public IP again and proxy [17:22:41] (03PS1) 10coren: Tool Labs: further tweaks of submit cron [operations/puppet] - 10https://gerrit.wikimedia.org/r/120563 [17:46:25] robh: can you please create a follow up ticket for 7102 ? [17:48:25] (03CR) 10coren: [C: 032] "Guaranteed* bug-free!" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120563 (owner: 10coren) [17:49:58] ottomata: how goes the migration of labs analytics? [17:51:55] hmm, andrewbogott, i think it is done [17:52:04] we shoudl confirm with dan and christian [17:52:08] but i think anything left can be scrapped [17:52:13] i can delete the nodes I'm responsible for [17:52:41] ottomata: can you please confirm and then update https://wikitech.wikimedia.org/wiki/Labs_Eqiad_Migration/Progress appropriately? [17:55:16] hrmm, damn it [17:55:32] matanya: if its not decommisioned yet then it doesnt get a wipe ticket [17:55:39] or someone will mistakenly make it [17:55:46] sure [17:55:47] andrewbogott: will do [17:55:53] spagewmf: Can you update me about your progress on the 'Editor Engagment' project? I'm unclear on whether it is done with migration or still needs work... [17:55:55] matanya: So if these werent decom'd yet what followup ticket do you mean? [17:56:00] (or did you mean for wipe?) [17:56:08] they were deccomed [17:56:14] and hence ticket is resolved [17:56:28] i requested a wipe ticket [17:56:33] ahh, ok, then its ok now [17:56:40] it said 'decom these' [17:56:44] and then no actual 'this is done' [17:56:52] see the change by mutante [17:57:06] he merged his commit there in the ticket [17:57:06] matanya: yea but there isnt a note in the ticket [17:57:17] he did? [17:57:26] what's up [17:57:29] oh, sorry. forgot to mention it, i guess [17:57:34] hi mutante 7102 [17:57:34] matanya: yes, you resolved it with no comment [17:57:36] thats not ok [17:57:40] ;p [17:57:49] * matanya facepalms [17:57:53] hehe [17:57:56] caught youuuuu [17:58:01] * robh teases in good fun [17:58:05] shame on me [17:58:13] * matanya is fired! [17:58:17] hehe, how dare you fuck up one in a thousand tickets, shame. [17:58:31] oh, you can't fire a volunteer [17:58:37] eh, so that ticket isn't done yet [17:58:42] indeed [17:58:44] yeah, my bad [17:58:49] the wiping ticket should exist but be linked to it [17:59:35] but be stalled or have a shit ton of references to say 'dont wipe thisuntil they are decom'd' [18:00:25] marktraceur: I have a bugzilla bug and also a project ('multimedia') that I'm hoping you will close out. Can you update https://bugzilla.wikimedia.org/show_bug.cgi?id=62616 and https://wikitech.wikimedia.org/wiki/Labs_Eqiad_Migration/Progress please? [18:00:35] yea, or use "depends on" when linking.. either way [18:01:13] marktraceur: Ah, looks like 'orgcharts' is your thing as well. [18:01:15] since we always have those 2 tickets for the same host, one in core and one in pmtpa that way, i started calling them "shutdown" vs. "decom" in some cases, to avoid cnfusio [18:01:19] ok, be back after meeting [18:01:45] so, robh please revert me :/ [18:02:19] andrewbogott: I cannot verify here - I'm on a plane now but I will deal with it later today [18:02:49] marktraceur: good enough, thank you! Once projects are migrated please move them into the 'Migration finished' section so I know I can stop nagging :) [18:03:27] matanya: robh , moving it back to core-ops, we'll continue after meeting [18:03:33] andrewbogott: I haven't started 'Editor Engagement' instance migration, I will do the named hosts this week. [18:03:57] spagewmf, can I do anything to help? This is way behind schedule [18:04:04] I can move instances intact if that's useful. [18:04:11] Well, possibly, depending on the instance. [18:07:06] andrewbogott: sorry. The instances I know about are relatively stock mediawiki-install and labs-vagrant instances. I'll read the instructions now. [18:08:20] spagewmf: ok, let me know if you need help with anything [18:22:43] !log springle synchronized wmf-config/db-eqiad.php 's5 db1037 full steam' [18:22:48] Logged the message, Master [18:26:12] varnish question: if I cachebust with a request parameter, does it create duplicates at any varnish level? or does it have no effect at all on the frontend/backend of varnish? [18:27:12] the reason I'm asking this is that we're considering the use of a request parameter for varnish to return an http header to tell the browser this will be a download (for a feature where users will be able click on a link to download a CDN-served image to save it) [18:27:47] I just want to make sure it won't result in any waste of resources [18:33:21] the headers show "miss, miss, miss" when I add a request parameter, which does suggest a level of waste [18:34:12] follow-up question would be whether varnish could be told to pretend that the request parameter isn't there for caching purposes [18:57:55] gwicke, I'm going to migrate those instances right now. [18:58:18] andrewbogott: great, thanks! [18:58:37] andrewbogott, wait a moment actually [18:58:48] I can delete some rt test data that we don't need any more [18:58:52] should speed up the migration [18:59:14] gwicke: um… the script is running already. Unless it's tb of data it should be fine. [18:59:27] it's around 50G [18:59:45] don't get a connection anyway, so nm [18:59:56] yeah, it's shut down for the sync [19:00:41] gwicke: looks like dump-api.wmflabs.org is pointing to nowhere, so I'm going to free it... [19:01:52] cmjohnson1: wanna move an19? [19:02:08] sure [19:02:15] k shutting ti down [19:02:17] andrewbogott, that was a https proxy front-end for what now lives at http://parsoid-tests.wikimedia.org/ [19:02:25] IIRC [19:02:26] ok [19:02:35] no https at the new location either [19:02:41] !log stopping hadoop services on analytics1019 and shutting it down for move to Row D [19:02:45] cmjohnson1: , you got dns change? [19:02:46] Logged the message, Master [19:02:53] umyes [19:02:56] yes [19:03:03] or maybe there is actually https there, wait.. [19:03:08] yup, works [19:03:09] ottomata: need help with users on stat1? [19:03:17] andrewbogott, so +1 for nuking [19:03:24] yes! [19:03:28] matanya: that would be amazing! [19:03:52] matanya: https://rt.wikimedia.org/Ticket/Display.html?id=6789 [19:04:21] PROBLEM - Host analytics1019 is DOWN: PING CRITICAL - Packet loss = 100% [19:05:12] oh, i guess apergos has done work on this too [19:05:21] matanya: , will you coordinate with apergos? [19:06:53] I think he knows bout that ticket and added a bunch of gerrit changesets actually [19:09:17] (03PS1) 10coren: Tool Labs: add postgresql-client [operations/puppet] - 10https://gerrit.wikimedia.org/r/120574 [19:09:32] (03PS3) 10Dzahn: decom: remove searchidx2 [operations/puppet] - 10https://gerrit.wikimedia.org/r/120145 [19:11:56] paravoid, do you have any objects to my new calculations at https://www.mediawiki.org/wiki/Talk:Requests_for_comment/Reducing_image_quality_for_mobile [19:13:28] (03PS1) 10Ottomata: Fixing Stefan's public key [operations/puppet] - 10https://gerrit.wikimedia.org/r/120575 [19:13:50] (03CR) 10Ottomata: [C: 032 V: 032] Fixing Stefan's public key [operations/puppet] - 10https://gerrit.wikimedia.org/r/120575 (owner: 10Ottomata) [19:14:05] (03PS1) 10Cmjohnson: updating dns entry for analytics1019 [operations/dns] - 10https://gerrit.wikimedia.org/r/120576 [19:14:50] (03CR) 10Cmjohnson: [C: 032] updating dns entry for analytics1019 [operations/dns] - 10https://gerrit.wikimedia.org/r/120576 (owner: 10Cmjohnson) [19:15:10] (03CR) 10Dzahn: [C: 032] decom: remove searchidx2 [operations/puppet] - 10https://gerrit.wikimedia.org/r/120145 (owner: 10Dzahn) [19:16:04] ottomata: yours to make if cfg changes [19:16:36] ottomata: i'll try to catch him tomorrow [19:19:52] !log searchidx2 - revoked puppet cert,remove from puppet,icinga,salt... [19:19:58] Logged the message, Master [19:22:17] ok thanks cmjohnson1, works great [19:22:18] on to an20? [19:22:22] ready? [19:22:26] sure [19:23:04] ye[ [19:23:06] yes [19:23:07] you can go ahead and change dns [19:23:30] !log stopping hadoop service on analytics1020 and shutting down for move to Row D [19:23:35] Logged the message, Master [19:24:29] (03PS1) 10Cmjohnson: updating dns for analytics1020 [operations/dns] - 10https://gerrit.wikimedia.org/r/120579 [19:25:01] PROBLEM - Host analytics1020 is DOWN: PING CRITICAL - Packet loss = 100% [19:25:12] (03CR) 10Cmjohnson: [C: 032] updating dns for analytics1020 [operations/dns] - 10https://gerrit.wikimedia.org/r/120579 (owner: 10Cmjohnson) [19:30:50] ottomata: booting now [19:36:21] !log shutdown searchidx2 [19:36:27] Logged the message, Master [19:37:25] (03PS1) 10DamianZaremba: Moving labs icinga from pmtpa to eqiad [operations/puppet] - 10https://gerrit.wikimedia.org/r/120581 [19:40:26] (03CR) 10Dzahn: [C: 032] Moving labs icinga from pmtpa to eqiad [operations/puppet] - 10https://gerrit.wikimedia.org/r/120581 (owner: 10DamianZaremba) [19:41:10] (03CR) 10Dzahn: [V: 032] Moving labs icinga from pmtpa to eqiad [operations/puppet] - 10https://gerrit.wikimedia.org/r/120581 (owner: 10DamianZaremba) [19:41:32] (03PS2) 10Ori.livneh: Geo-cookie: enable on one production text Varnish (cp1066) [operations/puppet] - 10https://gerrit.wikimedia.org/r/119098 [19:41:45] bblack: shall we start with that ^ ? i can deploy if you like [19:42:59] (03PS1) 10RobH: removing stat1002 access for amir & santhosh [operations/puppet] - 10https://gerrit.wikimedia.org/r/120583 [19:43:46] (03CR) 10jenkins-bot: [V: 04-1] removing stat1002 access for amir & santhosh [operations/puppet] - 10https://gerrit.wikimedia.org/r/120583 (owner: 10RobH) [19:43:58] PROBLEM - Puppet freshness on labstore2 is CRITICAL: Last successful Puppet run was Fri 21 Mar 2014 01:17:26 AM UTC [19:44:55] (03PS2) 10RobH: removing stat1002 access for amir & santhosh [operations/puppet] - 10https://gerrit.wikimedia.org/r/120583 [19:45:14] (03PS1) 10DamianZaremba: Updating NRPE allowed hosts to new icinga instance (pmtpa to eqiad migration) [operations/puppet] - 10https://gerrit.wikimedia.org/r/120584 [19:47:08] (03CR) 10coren: [C: 032] "Straightforward enough." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120584 (owner: 10DamianZaremba) [19:48:55] (03CR) 10RobH: [C: 032] removing stat1002 access for amir & santhosh [operations/puppet] - 10https://gerrit.wikimedia.org/r/120583 (owner: 10RobH) [19:49:03] (03CR) 10coren: [V: 032] "... why did jenkins-bot only +1?" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120584 (owner: 10DamianZaremba) [19:49:41] Damianz: im merging your change live cuz its mixed with mine on palladium Coren ^ [19:49:52] robh: I was about to tell you the same. [19:49:53] Coren: because Damianz is the author and not in trusted user regex [19:49:56] hehe [19:50:08] I love that feature of zuul [19:51:08] Ah. Interesting. That seems a little pointles since there is still need for a +2 review. [19:51:09] Damianz: $somebody should just add you:) [19:51:32] he is away [19:51:58] I really don't commit to puppet enough tbh [19:51:58] robh: when you are done with that, mind looking into https://rt.wikimedia.org/Ticket/Display.html?id=6133 and answer my question? [19:52:52] its mid migration last i checked, i'll find out in a bit [19:53:09] working access requests at the moment. [19:55:16] (03CR) 10Dzahn: [C: 032] "also not in Icinga anymore" [operations/dns] - 10https://gerrit.wikimedia.org/r/120062 (owner: 10Dzahn) [19:55:43] !log DNS update - removing tampa search pools and searchidx2 [19:55:50] Logged the message, Master [20:00:08] PROBLEM - HTTP 5xx req/min on tungsten is CRITICAL: CRITICAL: reqstats.5xx [crit=500.000000 [20:03:05] RECOVERY - Host analytics1019 is UP: PING OK - Packet loss = 0%, RTA = 0.27 ms [20:03:15] RECOVERY - Host analytics1020 is UP: PING OK - Packet loss = 0%, RTA = 0.40 ms [20:05:26] (03PS3) 10Ottomata: RT 7090 - adding analytics contact group to researchdb icinga alerts [operations/puppet] - 10https://gerrit.wikimedia.org/r/119797 [20:05:34] (03CR) 10Ottomata: [C: 032 V: 032] RT 7090 - adding analytics contact group to researchdb icinga alerts [operations/puppet] - 10https://gerrit.wikimedia.org/r/119797 (owner: 10Ottomata) [20:06:35] (03CR) 10coren: [C: 032] "Simple package addition." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120574 (owner: 10coren) [20:07:00] (03PS1) 10coren: Tool Labs: split HBA config from execnode [operations/puppet] - 10https://gerrit.wikimedia.org/r/120589 [20:07:54] !log deployed Parsoid fa03dd20 with deploy repo sha e4d28e7e [20:08:00] Logged the message, Master [20:08:03] (03PS2) 10coren: Tool Labs: split HBA config from execnode [operations/puppet] - 10https://gerrit.wikimedia.org/r/120589 [20:09:35] yay for a working sudo setup for parsoid deploys! [20:09:42] gwicke: I've finished moving the instances… http://parsoid.wmflabs.org/ doesn't look so good but I haven't investigated yet. [20:10:33] andrewbogott, the code is on /data/project [20:10:37] gwicke: well, actually, it looks like port 80 was never open for that instance, so… I don't know what that was for :) [20:10:39] in case that makes a difference [20:10:43] (03PS3) 10Dzahn: let yuvipanda upload mobile tarballs [operations/puppet] - 10https://gerrit.wikimedia.org/r/119336 [20:10:54] andrewbogott, we opened it in the firewall [20:11:12] gwicke: I can do that now, but it was definitely not open before. *shrug* [20:11:35] all ports were open before [20:11:55] PROBLEM - Packetloss_Average on oxygen is CRITICAL: packet_loss_average CRITICAL: 9.65246402062 [20:12:04] we are using a few other ports as well [20:12:15] (03CR) 10Dzahn: [C: 031] "updated per ticket comments, should have been Yuvi, and not mix 2 people in one change" [operations/puppet] - 10https://gerrit.wikimedia.org/r/119336 (owner: 10Dzahn) [20:12:15] ssh, another http port for the deb repo etc [20:12:28] but I can fix that too [20:13:33] omg, you're right, ports 1-65000 were open. [20:13:37] I would… not encourage that! [20:15:49] RECOVERY - Packetloss_Average on oxygen is OK: packet_loss_average OKAY: 1.44533255102 [20:15:57] gwicke: it's not obvious to me how this web server was set up… I encourage you to ssh in and see what you can see. [20:19:58] (03CR) 10coren: [C: 032] "What's the worse that could happen?" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120589 (owner: 10coren) [20:23:26] (03CR) 10RobH: [C: 04-1] "The linked RT ticket in question has raised questions to the need for this level of sudo access. My vote is indicated as a blocker until " [operations/puppet] - 10https://gerrit.wikimedia.org/r/119790 (owner: 10Ori.livneh) [20:27:41] gwicke: ? [20:28:10] andrewbogott, looking into it [20:28:14] ok, thanks [20:29:11] http://208.80.155.167:8000/ is working [20:30:29] gwicke: are other things not working? Or is that what it's supposed to do? [20:31:09] http://parsoid.wmflabs.org:8000/ does the same [20:31:48] andrewbogott, is DNS already updated? [20:31:53] it is for me [20:32:25] on my host it still point to the old location [20:32:30] *points [20:33:05] ok -- it'll catch up. [20:33:16] Can I mark this as 'finished' or are there other issues? [20:33:34] port 80 is working now too [20:33:53] andrewbogott, as far as I can tell things are fine [20:33:59] so thanks! [20:34:09] cool. Let me know if you run into any trouble. [20:37:25] hmm, so my wiki appears to have gone missing [20:37:53] (legalteamwiki) redirecting to wikimediafoundation.org ... it was working for the past couple weeks ... not seeing any apache changes that would do it.. [20:38:31] mark, do you have a minute to revew the archiva module? i really just want someone to tell me if the rsync server i'm setting up is ok [20:38:35] (03CR) 10BryanDavis: "I can't see the RT ticket, but my $0.02 is that sudo is needed on the boxes to restart the logstash server when it forgets what it's suppo" [operations/puppet] - 10https://gerrit.wikimedia.org/r/119790 (owner: 10Ori.livneh) [20:38:44] i can point you to what i'd like an opinion on [20:39:05] oh why does git-deploy never complete the fetches :( [20:39:18] andrewbogott_afk, I'll remove the tampa IP now; labs-ns0.wikimedia.org still returns both IPs which results in round-robin [20:53:09] (03PS1) 10Tpt: Activate the "other projects" sidebar managed by Wikibase in frwikisource [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120595 [20:55:26] (03CR) 10Lydia Pintscher: [C: 031] "Good from PM side :)" [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120595 (owner: 10Tpt) [20:56:35] Reedy: around? [20:56:36] robh: what is the fate of ES in tampa? es4,7,8 [20:56:50] (03CR) 10Hoo man: [C: 04-1] "Every configuration change should have a bug: Please open and and link it in the commit summary. Don't forget to note that you have commun" [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120595 (owner: 10Tpt) [20:57:21] bd808: you might know as well [20:57:43] matanya: What's up? [20:58:02] hi there, all good, hope you too :) [20:58:16] i'm wondering about tampa es servers [20:58:25] es4,7,8 [20:58:49] what do you mean by fate? [20:58:59] matanya ^ [20:59:08] moved to eqiad, decommed, recycle etc [20:59:16] searchidx2, just shut it down [20:59:24] matanya: No idea. [20:59:30] support contract litereally ends in 2 days [20:59:33] :) [20:59:54] searchidx2 is es servers ? [21:00:00] no [21:00:01] the es servers are relocating [21:00:35] any reason to keep them up now ? [21:01:09] (03PS1) 10RobH: Add csteipp & aaron to admins::jenkins group [operations/puppet] - 10https://gerrit.wikimedia.org/r/120596 [21:01:13] mutante: any ticket on those? [21:01:15] I think we have a decom ticket for es1-4 now...not sure about the other 2 [21:02:06] if there is, i don't see it linked to 6099 [21:02:07] matanya: 6266 [21:02:16] oh [21:02:19] es = external storage [21:02:25] external storage, not elastic search [21:02:29] andrewbogott_afk: I ran `sudo puppetd -tv` on ee-flow-extra (labs-vagrant role) has been running for an hour [21:02:54] matanya: yea, common confusion, we even had a bot trigger to explain because of that :p [21:02:57] !es [21:03:08] that is dead :) [21:03:14] yes :( [21:03:22] it knew all the server prefixes [21:03:29] so, not going yet [21:03:36] nope [21:03:53] thanks for clarifying this [21:04:05] 6266 getting an update would still be great! [21:04:15] i don't know much more than that [21:04:46] who is in charge of that? aaron ? [21:05:07] paravoid: Aye [21:05:09] (03PS2) 10John F. Lewis: Activate the "other projects" sidebar managed by Wikibase in frwikisource [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120595 (owner: 10Tpt) [21:05:36] (03CR) 10John F. Lewis: [C: 031] "Bug now attached, looks good." [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120595 (owner: 10Tpt) [21:05:41] Reedy: hey, we got an RT by James_F, legalteam wiki redirects to foundationwiki for some reason [21:06:02] (03CR) 10RobH: [C: 032] Add csteipp & aaron to admins::jenkins group [operations/puppet] - 10https://gerrit.wikimedia.org/r/120596 (owner: 10RobH) [21:06:05] I can't reproduce it on the appservers, any ideas on what would generate this? [21:06:16] I'm suspecting a stale appserver... [21:06:18] paravoid: *jamesofur :p [21:06:22] matanya: i think springle.. but .. [21:06:35] oh, sorry [21:06:49] paravoid: 10 [21:06:55] 10? [21:06:56] oops, ignore that [21:06:58] late night for him, will try to ask him tomorrow [21:07:11] But yeah - It keeps redirecting to foundationwiki. DNS, Apache and all that looks good but hey - I'm not a professional ops guy :p [21:07:26] mutante: if you want to get rid of hume: https://gerrit.wikimedia.org/r/#/c/74591/ is the last blocker [21:07:36] paravoid: Have we got an example URL? [21:07:38] paravoid: nothing, wrong usage of irc client :p [21:08:17] actully, i should point Reedy to that blocker, not you mutante [21:08:34] but yea, who moves the es boxes [21:08:47] Reedy: mind if i finish that one ? [21:09:08] ottomata: Yes, everything in labs that I care about got migrated. [21:09:08] sure [21:09:18] great [21:09:27] thanks Reedy [21:09:40] matanya: well, i got you some cleanup on maintenance.pp, rebase it [21:09:49] i saw [21:10:01] i rebase more than any other git command [21:12:12] found it [21:12:13] mw1163 [21:13:39] Reedy: I'm resyncing apache-config to mw1163; can you double-check that mediawiki is deployed as it should [21:13:43] also, this used to work... [21:13:53] boxes coming back after downtime and getting the new config [21:15:04] (03PS2) 10BBlack: GeoIP cookie: expand deployment from Labs to production [operations/puppet] - 10https://gerrit.wikimedia.org/r/119014 (owner: 10Ori.livneh) [21:16:57] paravoid: Reedy, used apache-fast-test with pybal option? [21:17:14] it actually checks all pooled servers and if they are all the same it summarizes [21:17:23] I did not, I used salt [21:17:25] but good to know, thanks [21:17:26] it's quickest to detect rogue apache [21:17:32] mutante: That's why I asked for an example url ;) [21:18:01] if you get only line as a result it means they were all identical [21:18:07] (I fixed it) [21:18:30] (03PS8) 10Matanya: Properly puppeti[sz]e purge-checkuser [operations/puppet] - 10https://gerrit.wikimedia.org/r/74591 (owner: 10Reedy) [21:18:37] Reedy: mutante ^ [21:19:26] arrg white space [21:20:05] (03CR) 10BBlack: [C: 032 V: 032] GeoIP cookie: expand deployment from Labs to production [operations/puppet] - 10https://gerrit.wikimedia.org/r/119014 (owner: 10Ori.livneh) [21:20:13] (03PS1) 10RobH: adding dumps.wikimedia.org certificate [operations/puppet] - 10https://gerrit.wikimedia.org/r/120598 [21:21:16] dumps is getting a cert? [21:21:36] (03PS9) 10Matanya: Properly puppeti[sz]e purge-checkuser [operations/puppet] - 10https://gerrit.wikimedia.org/r/74591 (owner: 10Reedy) [21:22:25] Reedy: apergos wants to support https so yep [21:22:45] (03CR) 10RobH: [C: 032] adding dumps.wikimedia.org certificate [operations/puppet] - 10https://gerrit.wikimedia.org/r/120598 (owner: 10RobH) [21:22:58] seems ready to me [21:23:33] did someone merge my change on palladium? [21:23:55] Sweet [21:24:00] considering lack of auth and pubicity of data published, the only thing someone can sniff is what dump someone is downloading. but that's also sniffable from the size of transfer:) [21:24:05] if someone will review this, i can go forward and decom hume. ops, please ? :) [21:24:15] apergos: dataset2 doesn't seem to redirect to dumps :( [21:24:15] http://dataset2.wikimedia.org/ [21:24:20] no [21:24:23] it should not [21:24:34] people shouldn't really use those hostnames for web access [21:24:38] how come? [21:24:39] lol [21:24:55] because we might be using one server or another [21:25:10] apergos: I you want help with 6789 i'll be glad to help [21:25:11] like, is it dataset2? dataset1001? the server at $newdc when it happens? [21:25:13] sorry [21:25:20] I was meaning.. [21:25:24] *if [21:25:37] why doesn't dataset2 -> dumps. not why shouldn't people be using dataset2.wm.o [21:25:42] Currently cleaning up [21:25:42] [21:25:52] well it should not redirect there [21:26:18] http://dataset2.wikimedia.org/backup-index.html [21:26:34] if someone really wants the files and they use that hostname, there they are [21:26:57] but in fact those files are 2 hours out of date since they are rsynced from the primary (now in eqiad) [21:27:12] right [21:27:14] what is this exclusion pattern? [21:27:17] so it's gotta stay for now :( [21:27:20] maybe that's what I'm missing [21:27:20] https everywhere [21:27:22] ah [21:27:27] ok so that's aother story [21:27:32] :) [21:27:40] likes anything that makes that regex shorter [21:27:41] dataset2, if someone ever uses it, yeah it's going to have a bad cert (one day) [21:27:46] it had a whole host (hah!) of NXDOMAIN [21:27:48] and dataset1001 the same [21:27:51] and anything that updates https://wikitech.wikimedia.org/wiki/Httpsless_domains [21:28:08] but dumps and download will sooon be https, what file is that in? I need to make a note to update that [21:28:19] upstream ? [21:28:21] ah [21:28:22] EFF [21:28:24] I'm just making an update to it now [21:28:24] yeah still [21:28:42] guess I should leave download|dumps|sitemap in for now... [21:28:43] if you have a link I'll update it when download becomes httpsable [21:29:16] and I'm too tired to even groan at the host pun so nyah [21:29:41] I make a commit in my fork on github and poke them on OFTC in #https-everywhere [21:29:44] [21:30:24] MaxSem: it's not that ooooh someone could sniff, it's sthat people want to turn on httpseverywhere and I've gotten complaints (there's ben an open bug for awhile) [21:30:59] someday there will be no http. and maybe that will make the internet a tiny bit safer [21:31:04] what's with "cs" and "cz" [21:31:13] or make the nsa spooks lose their hair a little sooner... [21:32:17] https://github.com/reedy/https-everywhere/tree/Wikimedia24032014 [21:32:21] I dunno, those are weird aren't they [21:32:45] ok reedy thanks for that [21:33:17] Feel free to ping/poke me to make another commit when it's done and I'll get it upstreamed [21:33:22] (03PS1) 10coren: Tool Labs: tweaks to xcrontab [operations/puppet] - 10https://gerrit.wikimedia.org/r/120600 [21:34:11] yay [21:34:41] Coren: icehouse next? [21:44:55] (03Abandoned) 10Dr0ptp4kt: WIP: DO NOT MERGE YET. Split up tagging for baselining period. [operations/puppet] - 10https://gerrit.wikimedia.org/r/119781 (owner: 10Dr0ptp4kt) [21:52:01] mutante: you can update yongle.wikimedia.org [21:52:09] it doesn't exist [21:53:50] update where? you meant the regex? [21:53:53] then it's Reedy [21:54:03] matanya: https://wikitech.wikimedia.org/w/index.php?title=Httpsless_domains&action=edit [21:54:05] https://wikitech.wikimedia.org/wiki/Httpsless_domains [21:54:31] Reedy: what is this??? a wiki? how do you edit it? [21:54:42] :P [21:54:47] Ctrl + A [21:54:48] Delete [21:54:50] Save [21:55:19] the wiki way of rm -rf / :) [21:55:40] done [21:55:41] help, no visual editor?:) [21:56:28] kinda, apart from the fact i used VE about 20 times total [21:57:47] Reedy: most of those redirects should go away [21:57:57] esp those pointing to tampa [21:58:59] matanya: the RT links in that wiki page, they all used to work [21:59:08] somebody deleted/broke template or something [21:59:44] used to [21:59:52] well i'll check those too [22:00:03] my pile is growing lately [22:01:10] or, just replace with BZ links [22:01:26] i'll review it in saner hours [22:01:35] (03CR) 10coren: [C: 032] Tool Labs: tweaks to xcrontab [operations/puppet] - 10https://gerrit.wikimedia.org/r/120600 (owner: 10coren) [22:02:54] matanya: move it to the very bottom of the list:) [22:03:51] how's stats.wm doing [22:03:59] i think i fixed that [22:04:06] some time ago [22:04:28] yep https://bugzilla.wikimedia.org/show_bug.cgi?id=32143 [22:07:44] cool [22:08:26] ok, mutante in your own time please https://gerrit.wikimedia.org/r/#/c/74591/ then i'll push hume decom patch [22:09:48] (03PS1) 10Dzahn: decom snapshot1-4 [operations/puppet] - 10https://gerrit.wikimedia.org/r/120615 [22:10:55] no RT linked in this ^ 7097 [22:11:42] (03PS2) 10Matanya: decom snapshot1-4 [operations/puppet] - 10https://gerrit.wikimedia.org/r/120615 (owner: 10Dzahn) [22:13:50] Reedy: purge_checkuser .so that should already exist, just wasnt puppetize. does that sound right to you [22:14:00] looks for it on terbium [22:14:27] i just see other "purge"things [22:14:47] oh, hume of course, i guess [22:14:53] that was the blocker mentioned earlier:) [22:15:59] (03PS1) 10Gilles: Add ?download parameter to images [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 [22:17:22] (03CR) 10Gilles: "This is untested because I don't know how to get varnish working on vagrant. If there's an easy way to set that up, please let me know." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [22:22:17] (03CR) 10Dzahn: [C: 031] "+1 for puppet code, but where is this actually running now and does it work?" [operations/puppet] - 10https://gerrit.wikimedia.org/r/74591 (owner: 10Reedy) [22:23:51] (03CR) 10Reedy: "I think it's a manual cronjob entry somewhere on hume..." [operations/puppet] - 10https://gerrit.wikimedia.org/r/74591 (owner: 10Reedy) [22:24:44] Reedy: can't find :p [22:24:47] (03PS1) 10coren: Tool Labs: Make xcrontab support @syntax [operations/puppet] - 10https://gerrit.wikimedia.org/r/120682 [22:28:47] (03CR) 10Dzahn: "i can't seem to find it on hume, which user? tried apache,root,mwdeploy , /etc/cron.d/ ..." [operations/puppet] - 10https://gerrit.wikimedia.org/r/74591 (owner: 10Reedy) [22:29:44] robh: thanks for doing some of those mailing list bugs [22:29:48] (03CR) 10Hoo man: [C: 04-1] Add ?download parameter to images (032 comments) [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [22:30:04] Jamesofur: im looping though them all, quite welcome [22:30:13] robh: If I knew it was you - I would have thanked you on IRC :p [22:32:48] (03CR) 10coren: [C: 032] Tool Labs: Make xcrontab support @syntax [operations/puppet] - 10https://gerrit.wikimedia.org/r/120682 (owner: 10coren) [22:34:51] heh [22:40:59] (03PS1) 10Andrew Bogott: Move the labs ganglia aggregator to eqiad. [operations/puppet] - 10https://gerrit.wikimedia.org/r/120687 [22:41:01] when someone asks for a list, and we ask followup on what to call it [22:41:07] and it sits since last july [22:41:11] im just rejecting it... [22:43:19] (03CR) 10Andrew Bogott: [C: 032] Move the labs ganglia aggregator to eqiad. [operations/puppet] - 10https://gerrit.wikimedia.org/r/120687 (owner: 10Andrew Bogott) [22:44:16] PROBLEM - Puppet freshness on labstore2 is CRITICAL: Last successful Puppet run was Fri 21 Mar 2014 01:17:26 AM UTC [22:47:31] 89 fixed-address statistics1.wikimedia.org; [22:47:33] heh [22:49:42] (03PS1) 10Dzahn: decom ms5 [operations/puppet] - 10https://gerrit.wikimedia.org/r/120689 [22:53:15] (03PS2) 10Gilles: Add ?download parameter to images [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 [22:53:42] (03CR) 10Gilles: Add ?download parameter to images (032 comments) [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [22:54:31] (03PS1) 10Dzahn: remove "statistics1" from DHCP [operations/puppet] - 10https://gerrit.wikimedia.org/r/120694 [22:57:50] !log Forced /srv/scap to update to c771a46 across the cluster [22:57:56] Logged the message, Master [22:58:37] bd808: Are you doing the SWAT today? [22:59:28] we have things to deploy [22:59:36] (03CR) 10Hoo man: [C: 04-1] "Now you set headers on the request object, not the response (that's not what you want). Also you will probably produce invalid request url" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [22:59:43] RoanKattouw: No. Just pushing a fix for a bug Antoine found in the scap code itself [23:00:00] making patch to update wmf19 submodule [23:00:12] bd808: OK. You know that maaaybe you shouldn't do that during a scheduled deployment window? [23:00:13] RoanKattouw: I commited a trivial PHP fatal error. [23:00:18] hah whoops [23:00:21] RoanKattouw: should have left bd808 to +2 the change though [23:00:41] RoanKattouw: double whops is that Jenkins only runs php -l on files ending with .php [23:00:55] RoanKattouw: You're right; I appologize [23:00:57] !quip [23:01:13] bd808: It's OK, it's just Yet Another Reason why we need a locking system [23:01:26] * bd808 nods [23:01:59] (03CR) 10Dzahn: [C: 032] "whatever this was, it's not stat1" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120694 (owner: 10Dzahn) [23:02:37] Short of blocking ssh to tin it's going to be pretty hard to accomplish [23:02:51] wrong approach :P [23:03:09] Even with trebuchet something like what I just did would work fine [23:03:35] You could hack that using git hook (although these atm don't even seem to work properly) [23:03:42] * a git hook [23:03:57] vim [23:04:15] A git hook won't stop modifications to the working copy which is what scap operates on [23:04:30] that, in more words, than my simple: "vim" [23:04:52] bd808: Right, but no one is supposed to do that, right [23:05:22] trebuchet only stops `git deploy start` with it's lock [23:05:39] OK so who's doing the SWAT? [23:05:56] https://gerrit.wikimedia.org/r/#/c/120697/ [23:06:07] i don't know how this works [23:06:10] I don't crea... I can do it [23:06:27] do we deploy or does a swat deployer do it? [23:06:44] a swatter [23:06:57] ori: ebernhardson ^ ? [23:06:58] hoo: The answer to your question is … complicated. Train deploys currently do make "live" changes on tin before adding them to gerrit and reverting the local changes to tin. [23:07:06] gerrit doesnt merge things for me.. did i miss out? [23:07:26] greg-g: ok [23:07:31] bd808: Yeah... there are edge cases... also recreating the iw-map and stuff [23:07:49] (03CR) 10Dzahn: [V: 032] remove "statistics1" from DHCP [operations/puppet] - 10https://gerrit.wikimedia.org/r/120694 (owner: 10Dzahn) [23:07:50] ori: ebernhardson now with content in my ping: time for a SWAT? [23:09:17] Submitted, Merge Pending [23:09:35] am I needed to deploy something? [23:09:45] ori: yup, please do us [23:09:50] https://gerrit.wikimedia.org/r/#/c/120697/ [23:09:51] let's see if you can merge [23:09:53] * our stuff [23:09:55] i cant [23:09:58] but other repo [23:10:11] it's for test.wikidata only [23:10:22] greg-g: ok with you? [23:10:51] yeah [23:11:32] cool. +2'd, gerrit running gate-and-submit. [23:11:40] :) [23:12:58] (03CR) 10Dzahn: "still "Submitted, Merge Pending".." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120694 (owner: 10Dzahn) [23:13:27] bblack: whoa, i completely missed the deployment [23:13:28] thanks! [23:13:29] TimStarling: you haven't filled up the disk on gallium or anything weird like that have you? [23:13:35] :) [23:14:12] I haven't filled up the disk, hopefully I haven't done anything weird [23:14:22] I just ran some tests [23:14:49] k, we'll give it a minute, just mutante has been having issues with jenkins in the last 15ish minutes [23:15:11] there it goes [23:15:24] (03CR) 10Gilles: "The requests will be coming from Media Viewer, where it will be just "?download" and no other parameter. "=1" won't happen, it's not neces" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [23:16:01] greg-g: TimStarling "16:15 < hashar> mutante: that is Gerrit JGit backend having trouble/delay merging a commit apparently [23:16:52] * greg-g shrugs [23:16:59] greg-g: ignore me, my bad [23:17:09] that one had a dependency it should not have had [23:17:12] well, I may possibly be doing bad things to it [23:17:22] !log ori synchronized php-1.23wmf19/extensions/Wikidata 'I6ffb304d3: Update Wikidata' [23:17:26] TimStarling: yeah, where my mind went... [23:17:28] Logged the message, Master [23:17:30] like writing to a shared sqlite database when I shouldn't be [23:17:34] ^ aude [23:17:37] yay! [23:17:47] * aude checks test wikidata [23:17:48] ok [23:17:52] hoo, aude: can you...yeah [23:17:54] thanks :) [23:18:44] gah, issue in PageTranslationHooks [23:18:53] i'll poke around more [23:18:57] what? [23:18:58] * greg-g heads out [23:19:10] give me a minute [23:19:16] oh [23:19:23] greg-g: should I sync the fix-up? [23:19:25] I don't mind [23:19:49] ori: yes, please [23:19:54] thanks [23:20:04] OK. make your train! :P [23:20:10] :) [23:20:12] the train-train I mean, not the dpeloyment train! [23:20:15] (bus) [23:20:22] deployment bus? [23:20:28] not the same ring to it [23:20:37] * greg-g gos [23:20:39] +e [23:22:02] might make another submodule update [23:22:30] ori: So are you today's SWATter? [23:23:06] i guess so, wouldn't mind splitting the job tho ;) [23:23:16] OK [23:23:25] Did you do all the Wikibase stuff already? [23:23:28] Then I can just do the VE stuff [23:24:09] (03PS1) 10Ori.livneh: CentralNotice: Re-set $wgCentralGeoScriptURL to false [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120699 [23:24:13] aude: Do you have the ext. locally? [23:24:27] RoanKattouw: wikidata needs to follow-up with a fix-up [23:24:29] OK [23:24:31] still waiting for it [23:24:44] ori: Let me know when is a good time to start doing the VE stuff [23:24:50] hoo: i don't but think i see the issue [23:25:30] (03CR) 10Ori.livneh: [C: 032] CentralNotice: Re-set $wgCentralGeoScriptURL to false [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120699 (owner: 10Ori.livneh) [23:25:40] (03Merged) 10jenkins-bot: CentralNotice: Re-set $wgCentralGeoScriptURL to false [operations/mediawiki-config] - 10https://gerrit.wikimedia.org/r/120699 (owner: 10Ori.livneh) [23:25:50] !log ori updated /a/common to {{Gerrit|I42e9c8c97}}: CentralNotice: Re-set $wgCentralGeoScriptURL to false [23:25:55] aude: Has that been there before? [23:25:56] Logged the message, Master [23:26:18] hoo: ? [23:26:25] !log ori synchronized wmf-config/CommonSettings.php 'I42e9c8c97: CentralNotice: Re-set $wgCentralGeoScriptURL to false' [23:26:31] Logged the message, Master [23:26:50] aude: I mean the fatal, have you checked test before [23:27:06] I wasn't on it today [23:27:10] it's new and have a paptch for it [23:27:16] patch* [23:27:34] no idea how i can't reproduce these issues on my test wiki, though don't have translate [23:28:08] even on beta wikidata [23:28:16] aude: would it make sense to have Roan proceed with the VE swat, and we'll ask to interject with your fix-up if/when it's ready? [23:28:18] no issue [23:28:26] ori: go ahead [23:28:29] RoanKattouw: ^ [23:28:34] would take us time to make a new build [23:28:37] (03CR) 10Dzahn: [C: 032] "that mount was already ensured absent and did not exist on hume, which is the only node using nfs::upload" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120689 (owner: 10Dzahn) [23:33:29] (03CR) 10Hoo man: "> The requests will be coming from Media Viewer, where it will be just "?download" and no other parameter. "=1" won't happen, it's not nec" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [23:34:33] RoanKattouw: you can go ahead, in case you missed my earlier message [23:35:07] as always, takes time to wait for jenkins etc. so go ahead [23:39:35] (03PS1) 10Dzahn: decom - remove ms5 [operations/dns] - 10https://gerrit.wikimedia.org/r/120705 [23:40:25] (03CR) 10Gilles: "Yes, people will find out about those URLs. That's why I'm being cautious, for example not settings the response header based on the custo" [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [23:40:27] (03CR) 10Dzahn: [C: 031] "removed from puppet in I8530218e320364849d1ae1ccfd1cea7a18489e90" [operations/dns] - 10https://gerrit.wikimedia.org/r/120705 (owner: 10Dzahn) [23:42:18] (03CR) 10Gilles: "Re-reading your remarks, you might have missed the fact that I don't clean "=bla" because it doesn't get through, the only recognized URL " [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [23:49:33] (03CR) 10Hoo man: "Yikes... that's what I get for only skimming over this change, sorry for the oversights." [operations/puppet] - 10https://gerrit.wikimedia.org/r/120617 (owner: 10Gilles) [23:49:40] aude: I have a hard stop in ten minutes [23:49:46] ok [23:49:53] we will be there by then [23:49:54] updating submodule now [23:50:16] ori: Crap, sorry [23:50:36] I went upstairs to fix a travel problem and forgot all about the SWAT thing [23:50:38] I'll do it now [23:50:41] RoanKattouw: no worries at all [23:51:38] !log Switched CentralNotice to GeoIP cookie rather than bits.wm.o/geoiplookup script tags [23:51:43] Logged the message, Master [23:52:37] ori: https://gerrit.wikimedia.org/r/#/c/120710/ [23:52:39] ori, RoanKattouw: https://gerrit.wikimedia.org/r/120710 [23:52:43] :) [23:53:16] Coren: Is the very slow speed of Beta Labs a more general issue, and does it have an end in sight? Was about to launch a round of user testing… [23:53:25] hoo: Ok I'll include that one [23:53:57] James_F: As far as I know, everything in labs is working fine. This is the first I hear of it. [23:54:07] Coren: Ah. Oh dear. :-( [23:54:25] Coren: Yeah, it's been near-catastrophic for the past few hours. [23:54:53] Coren: E.g. beta-bits ERR_CONNECTION_REFUSED, exceedingly slow loads, and all that fun. [23:55:01] Coren: Sorry, I assumed it was known. :-( [23:56:02] RoanKattouw: OK, seems like you're taking over that one (which is great for me), so I'm signing off [23:56:09] unless you'd like me to stick around [23:56:13] thanks ori [23:56:24] ori: Yup I got it [23:56:30] A quick look at other labs projects doesn't reveal any performance issue. It's something on beta afaict. Have you talked to Hashar? I know he's busy doing migration, there might be something he's doing that's consuming resources? [23:56:30] thanks, *wave* [23:56:34] Sorry again for the unexplained absence [23:57:14] !log ms5 - removed from icinga,puppet,storedconfigs,salt... [23:57:20] Logged the message, Master [23:57:51] Coren: I didn't ask him before he disconnected for the night, no. [23:58:19] I seem to remember hearing him talk about a database dump; perhaps that's what is going on? [23:58:35] Ah. [23:58:41] Coren, staging.wmflabs.org is also slow as hell