[00:04:37] MaxSem: :P [00:05:04] I'm serious - want to accelerate stuff a bit [00:05:29] and volunteer to be a scapegoat! [00:13:55] 3: Rethink #Blocked-on-MediaWiki-Core - https://phabricator.wikimedia.org/T85302#1016492 (10Krenair) [00:15:37] MaxSem: no, not today [00:15:54] MaxSem: but thank you :) [00:15:56] 3: Rethink #Blocked-on-MediaWiki-Core - https://phabricator.wikimedia.org/T85302#1016510 (10Legoktm) Would renaming the tag to "Blocked-on-MWCore-Team" help? [00:16:45] 3MediaWiki-Authentication-and-authorization, Project-Creators, MediaWiki-Core-Team: Create "MediaWiki-Authentication-and-authorization" component - https://phabricator.wikimedia.org/T88317#1016514 (10Legoktm) 5Open>3Resolved a:3Legoktm Created #authn. [00:17:45] 3MediaWiki-Authentication-and-authorization, MediaWiki-Core-Team: Document current MediaWiki PHP authn stack - https://phabricator.wikimedia.org/T88195#1016521 (10Legoktm) [00:32:21] 3MediaWiki-Core-Team: Rethink #Blocked-on-MediaWiki-Core - https://phabricator.wikimedia.org/T85302#1016595 (10bd808) >>! In T85302#1016510, @Legoktm wrote: > Would renaming the tag to "Blocked-on-MWCore-Team" help? I'd be cool with that. Renaming our #mwcore-team tag to `MWCore-Team` wouldn't bother me either.... [00:38:02] 3MediaWiki-Core-Team: Make a ReplicatedBagOStuff class for multi-DC use - https://phabricator.wikimedia.org/T88634#1016617 (10aaron) 3NEW a:3aaron [00:39:01] 3MediaWiki-Core-Team, MediaWiki-Configuration, Continuous-Integration: Update jenkins for extension registration changes - https://phabricator.wikimedia.org/T86359#1016626 (10Legoktm) [00:39:34] 3MediaWiki-Core-Team: ObjectCacheSessionHandler should avoid pointless writes in write() - https://phabricator.wikimedia.org/T88635#1016631 (10aaron) 3NEW a:3aaron [00:40:38] 3MediaWiki-Core-Team, operations: Store unsampled API and XFF logs - https://phabricator.wikimedia.org/T88393#1016638 (10tstarling) I think the reason rotation is done daily is because logrotate is a daily cron job, and does not support a shorter rotation period. It could be replaced by 50 lines of your favourit... [00:56:54] 3MediaWiki-Core-Team, operations: Store unsampled API and XFF logs - https://phabricator.wikimedia.org/T88393#1016697 (10ori) >>! In T88393#1016638, @tstarling wrote: > I think the reason rotation is done daily is because logrotate is a daily cron job, and does not support a shorter rotation period. It could be... [00:57:11] legoktm: How do you feel about turning your +1 into a +2 on https://gerrit.wikimedia.org/r/#/c/186897/ ? [00:58:27] :) [00:58:29] I was hoping AaronS would review it [00:59:39] I was too. I should have poked him I guess [01:10:58] ori: you recall the "Services report card" email you sent back in November? I'd like to copy/paste that into a phab task [01:11:38] robla: sure, go for it [01:11:42] cool, thanks! [01:13:48] csteipp: https://gerrit.wikimedia.org/r/#/c/188721/ [01:13:50] robla: http://www.thoughtworks.com/radar/techniques/microservice-envy [01:13:58] from the past week! [01:17:30] legoktm: https://gerrit.wikimedia.org/r/#/c/185372/ [01:19:54] bd808: https://gerrit.wikimedia.org/r/#/c/185243/ simple cleanups [01:23:46] 3MediaWiki-Core-Team, operations: move misc mw maintenance scripts into mw puppet module - https://phabricator.wikimedia.org/T88597#1016814 (10Dzahn) p:5Triage>3Normal [01:25:48] * ori thinks we should replace our monolithic, ball-of-mud app with a storage, authentication, monitoring, service discovery, http api service. [01:26:34] tough crowd! *stops trolling*. [01:26:38] I happen to like our monolith. It's like a monument to history :) [01:27:08] http://fc07.deviantart.net/fs71/f/2012/034/1/b/monolith_by_reneaigner-d4ojqar.jpg looks nice [01:28:43] Or http://th05.deviantart.net/fs70/PRE/i/2013/065/4/1/monolith_by_warrencarr-d5x6apb.jpg [01:29:27] Also, this is a really freaking cool table. http://www.yatzer.com/The-monolith-dining-table-by-Gioia-Meller-Marcovicz [01:30:08] that looks...sharp [01:30:36] with that table it's easy to imagine one of your guests ending up in the ER after a glass or two of champagne [01:30:57] ori "stops trolling" ...ha....haha.....hahahahahahaha :-) [01:31:18] "The monolith was made for a private collector in Paris, where it is now placed." [01:31:19] i didn't specify a duration [01:31:21] Awww, not for sale [01:32:30] I wish I had a picture of our conference table from Mediaphex. It was scary. Glass top and 1" square iron bar legs in a haystack sort of pattern. You looked down through the glass top on very shiny, very pointy metal [01:34:34] and there was a wall hanging behind it that clients affectionally called "the dead babies picture". Our whole office was very Trent Reznor inspired [01:38:14] legoktm: https://gerrit.wikimedia.org/r/#/c/186946/ [01:40:25] ori: https://graphite.wikimedia.org/render?from=-12hours&until=now&width=900&height=400&target=cactiStyle(color(alias(frontend.navtiming.mediaWikiLoadComplete.desktop.overall.mean%2C%22mean%22)%2C%22blue%22))&title=frontend.navtiming.mediaWikiLoadComplete.desktop.overall.mean [01:40:33] kind of back to normal [01:40:41] still noisier [01:42:01] * ori nods [01:42:45] AaronSchulz: left a comment [01:46:10] legoktm: it's funny when merge/cas was added I was very clear that there should be just one method that always works rather than one two (one that might blow up) [01:46:18] * AaronSchulz remembers talking to terry about that [01:46:57] legoktm: I already talked to the Flow people about multi-DC stuff, and how merge/cas suck for that [01:47:10] I think they plan on ripping it out anyway, so we can wait [02:03:40] legoktm: https://gerrit.wikimedia.org/r/#/c/186914/ [02:08:33] * legoktm waits for his grep to finish... [04:04:27] 3MediaWiki-Core-Team, MediaWiki-General-or-Unknown, Librarization: [Regression] MediaWiki should detect absent or outdated vendor - https://phabricator.wikimedia.org/T74777#1017020 (10bd808) Time to re-re-close this? I'm not sure what else I can reasonably do. [06:57:34] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1017167 (10Smalyshev) Preliminary notes: - Graph is a set of vertex collections and edge collections, both document collection with some special attributes. - ArangoDB keeps index... [07:16:27] 3MediaWiki-Core-Team, wikidata-query-service: Find another TinkerPop 3 implementation to add to the Wikidata Query candidate list - https://phabricator.wikimedia.org/T88551#1017195 (10Smalyshev) Here is the list of all TinkerPop implementations (not TP3, but may be helpful as a starting point): https://github.co... [09:45:45] 3MediaWiki-Core-Team, Wikimedia-General-or-Unknown: Existed pages without ability to reach and obviously wrong namespace - https://phabricator.wikimedia.org/T87645#1017430 (10Redrose64) I see several changes/patches/merges above, presumably these are code fixes: do we know when the data will also be fixed? I am... [10:05:56] 3MediaWiki-Core-Team: Improve performance of Captcha extension on edit forms - https://phabricator.wikimedia.org/T88661#1017439 (10aaron) 3NEW [10:27:11] 3MediaWiki-Core-Team: Multi-DC documentation RFC - https://phabricator.wikimedia.org/T88666#1017506 (10aaron) 3NEW a:3aaron [10:33:35] 3MediaWiki-Core-Team: MediaWiki multi-datacenter investigation and work - https://phabricator.wikimedia.org/T88445#1017515 (10aaron) [11:37:24] 3MediaWiki-Core-Team, wikidata-query-service: Figure out if Neo4j is a possible alternative to Titan - https://phabricator.wikimedia.org/T88571#1017609 (10daniel) Neo4j river plugin for Elastic: https://github.com/sksamuel/elasticsearch-river-neo4j [12:32:41] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1017722 (10Manybubbles) Looks like its out! Thanks for the report! Can you put it on the wiki page and grab the next one on the list? Tinkerpop 3 supports Neo4j so that might be qui... [13:06:40] 3MediaWiki-Core-Team: Multi-DC documentation RFC - https://phabricator.wikimedia.org/T88666#1017828 (10Aklapper) Note that there is a MediaWiki-RfCs project and a documentation one. (If that makes sense, don't know, description of this task is short and a bit cryptic. :) [13:12:06] 3MediaWiki-Core-Team, Wikimedia-General-or-Unknown: Existed pages without ability to reach and obviously wrong namespace - https://phabricator.wikimedia.org/T87645#1017842 (10Krenair) [13:23:11] 3MediaWiki-Core-Team, Wikimedia-General-or-Unknown: Existed pages without ability to reach and obviously wrong namespace - https://phabricator.wikimedia.org/T87645#1017854 (10Krenair) @tstarling uploaded https://gerrit.wikimedia.org/r/#/c/188748/ to improve the script we use to find/fix these. [15:08:35] 3MediaWiki-Core-Team, wikidata-query-service: Figure out if Neo4j is a possible alternative to Titan - https://phabricator.wikimedia.org/T88571#1017973 (10JanZerebecki) master/slave also means in this case no sharding. And it probably also means the work on one query will not be distributed over more than one no... [15:26:32] 3MediaWiki-Core-Team, Wikimedia-General-or-Unknown: Existed pages without ability to reach and obviously wrong namespace - https://phabricator.wikimedia.org/T87645#1018005 (10Rillke) Is it save to move these pages using API and the `fromid` parameter? In this case, let me fix Commons. Or tell me where you moved... [15:28:18] 3MediaWiki-Core-Team, wikidata-query-service: Figure out if Neo4j is a possible alternative to Titan - https://phabricator.wikimedia.org/T88571#1018006 (10Manybubbles) >>! In T88571#1017973, @JanZerebecki wrote: > master/slave also means in this case no sharding. And it probably also means the work on one query... [15:33:46] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1018010 (10Manybubbles) I just updated [[https://docs.google.com/a/wikimedia.org/spreadsheets/d/1MXikljoSUVP77w7JKf9EXN40OB-ZkMqT8Y5b2NYVKbU/edit#gid=0 | our spreadsheet]] with notes a... [16:29:33] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1018178 (10Fceller) Hi, I'm the CTO of ArangoDB, so my comments are most certainly biased. I still would like to tell you about our opinions on the raised issues, namely full-text inde... [16:33:26] innnnteresting [17:00:21] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1018246 (10Smalyshev) @Fceller many thanks to you for your explanations! I certainly understand the reasons for your choices, however I'm still not sure how given our data sizes - ~1... [17:02:13] <_joe_> interesting indeed [17:27:03] * csteipp gives greg-g a disappointed look: https://www.ssllabs.com/ssltest/analyze.html?d=creativecommons.org [17:30:30] csteipp: :/ [17:41:11] bd808: we just got off the monthly Wikidata call. manybubbles brought up the idea of an epic centered on data dumps that he'd like to pitch at some point. [17:41:37] just about to jump in a hangout to talk to him about that! :) [17:57:58] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1018425 (10Fceller) @Smalyshev May I ask how much main memory your server has? Also, can you elaborate I little bit, what you mean by "2-3 thousands indexed fields"? Are you going to... [18:02:27] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1018433 (10Smalyshev) @Fceller this was tested on my work machine, so it has about 8G of memory, are you saying giving it more memory would significantly improve things significantly?... [18:14:27] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1018451 (10Fceller) > There are thousands of indexable attributes, but not each document has every one of them - typical one would have tens, rarely up to a hundred. We need to be abl... [18:15:36] manybubbles: can't edit your spreadsheet :/ [18:15:45] probably not necessary though [18:27:39] thanks [18:32:15] bd808: I did make a patch to scap that jenkins -2'd because of 80col which is why I made the tox.ini patch (https://gerrit.wikimedia.org/r/#/c/188798/) [18:32:39] err [18:32:40] -1'd? [18:34:30] YuviPanda: *nod* but you just need a newline after the first comma in line 445 :) [18:34:45] I know but 80!!!!!!! [18:34:53] kids these days [18:35:18] bd808: I also realized that I've become an impatient merger, and that's a bad habit I'm picking up from operations/puppet and should change my ways. [18:35:23] When I was a boy we had 80x24 displays and we liked them [18:35:30] sorry about just merging, I facepalmed just as I hit submit [18:35:44] no worries. [18:37:33] bd808: :) 0 cherrypicks now on deployment-prep, btw :) And mw / jobrunner running prod roles, and no tiny /var on new hosts! [18:37:36] progresss! [18:37:48] massive! good work [19:12:53] 3MediaWiki-Core-Team, wikidata-query-service: Mourn Titan - https://phabricator.wikimedia.org/T88550#1018696 (10MaxSem) 5Open>3Resolved a:3MaxSem One day of mourning is enough. GET BACK TO WORK! :P [19:13:16] 3MediaWiki-Core-Team, wikidata-query-service: Mourn Titan - https://phabricator.wikimedia.org/T88550#1018699 (10Manybubbles) You can still play the 10 hours of sad violin if you really want though. [19:14:03] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1018710 (10Smalyshev) @Fceller for the test, I used two of the attributes that are present in each document for hash indexes, and two that may or may not be present, for skiplist ones.... [19:14:08] 3MediaWiki-Core-Team, wikidata-query-service: Investigate hand building a gremlin implementation against Elasticsearch for wikidata query - https://phabricator.wikimedia.org/T88577#1015218 (10Manybubbles) Updated https://docs.google.com/a/wikimedia.org/spreadsheets/d/1MXikljoSUVP77w7JKf9EXN40OB-ZkMqT8Y5b2NYVKbU/... [19:18:57] 3MediaWiki-Core-Team: Investigate Apache Accumulo - https://phabricator.wikimedia.org/T88714#1018730 (10Manybubbles) 3NEW [19:21:04] 3MediaWiki-Core-Team, wikidata-query-service: Find another TinkerPop 3 implementation to add to the Wikidata Query candidate list - https://phabricator.wikimedia.org/T88551#1018742 (10Manybubbles) I talked to the OrientDB folks and they are excited about the direction of TinkerPop3 but haven't started work on im... [19:21:59] 3MediaWiki-Core-Team: Investigate Bitsy for single node users - assuming it supports TinkerPop 3 and we're still using TinkerPop 3 - https://phabricator.wikimedia.org/T88715#1018743 (10Manybubbles) 3NEW [19:22:16] 3MediaWiki-Core-Team: Investigate Apache Accumulo - https://phabricator.wikimedia.org/T88714#1018752 (10Manybubbles) p:5Triage>3Normal [19:23:02] 3MediaWiki-Core-Team: Investigate BigData - https://phabricator.wikimedia.org/T88717#1018761 (10Manybubbles) 3NEW [19:23:14] 3MediaWiki-Core-Team, wikidata-query-service: Investigate BigData - https://phabricator.wikimedia.org/T88717#1018769 (10Manybubbles) p:5Triage>3Normal [19:23:21] 3MediaWiki-Core-Team, wikidata-query-service: Investigate Bitsy for single node users - assuming it supports TinkerPop 3 and we're still using TinkerPop 3 - https://phabricator.wikimedia.org/T88715#1018772 (10Manybubbles) [19:23:26] BigData sounds like a great name [19:23:31] 3MediaWiki-Core-Team, wikidata-query-service: Investigate Apache Accumulo - https://phabricator.wikimedia.org/T88714#1018730 (10Manybubbles) [19:24:04] 3MediaWiki-Core-Team, wikidata-query-service, Wikidata: Wikidata Query - add license - https://phabricator.wikimedia.org/T86833#1018784 (10Manybubbles) 5Open>3Resolved [19:24:16] 3MediaWiki-Core-Team, wikidata-query-service, Wikidata: Evaluate Titan as graph storage/query engine for Wikidata Query service - https://phabricator.wikimedia.org/T76373#1018787 (10Manybubbles) [19:24:17] 3MediaWiki-Core-Team, wikidata-query-service, Wikidata: Figure out quantity representation - https://phabricator.wikimedia.org/T85298#1018786 (10Manybubbles) 5Open>3Resolved [19:24:23] 3MediaWiki-Core-Team, wikidata-query-service: Investigate hand building a gremlin implementation against Elasticsearch for wikidata query - https://phabricator.wikimedia.org/T88577#1018788 (10Manybubbles) 5Open>3Resolved [19:24:32] 3MediaWiki-Core-Team, wikidata-query-service: Find another TinkerPop 3 implementation to add to the Wikidata Query candidate list - https://phabricator.wikimedia.org/T88551#1018789 (10Manybubbles) 5Open>3Resolved [19:28:28] 3MediaWiki-Core-Team, wikidata-query-service: Investigate InfoGrid - https://phabricator.wikimedia.org/T88721#1018810 (10Manybubbles) 3NEW [19:28:35] 3MediaWiki-Core-Team, wikidata-query-service: Investigate InfoGrid - https://phabricator.wikimedia.org/T88721#1018828 (10Manybubbles) p:5Triage>3Normal [19:28:43] 3MediaWiki-Core-Team, wikidata-query-service: Investigate InfoGrid for WDQ - https://phabricator.wikimedia.org/T88721#1018830 (10Manybubbles) [19:28:49] 3MediaWiki-Core-Team, wikidata-query-service: Investigate BigData for WDQ - https://phabricator.wikimedia.org/T88717#1018831 (10Manybubbles) [19:28:54] 3MediaWiki-Core-Team, wikidata-query-service: Investigate Apache Accumulo for WDQ - https://phabricator.wikimedia.org/T88714#1018832 (10Manybubbles) [19:29:06] 3MediaWiki-Core-Team, wikidata-query-service: WDQ: Investigate Bitsy for single node users - assuming it supports TinkerPop 3 and we're still using TinkerPop 3 - https://phabricator.wikimedia.org/T88715#1018833 (10Manybubbles) [19:32:36] titan is 750% java! https://www.openhub.net/p/titandb [19:33:48] <^d> lol [19:34:42] Who owns the wmgUseMonologLogger live hack? [19:34:50] <^d> ori. [19:34:57] <^d> Should be a patch merged for it? [19:34:59] is it deployed? [19:35:07] yeah, I just created https://gerrit.wikimedia.org/r/188861 [19:35:09] I'll merge it [19:35:11] if you're referring to the addition of 'false', that's fine [19:35:22] <^d> Oh that. [19:35:26] <^d> Different, nvm [19:35:27] Reedy: no need [19:35:35] no need? [19:35:44] I can't pull unless I stash it etc ;) [19:35:44] <^d> manybubbles: Also 60% ruby, 49% shell scripts, 1% other...for a total of 110% [19:35:47] because https://gerrit.wikimedia.org/r/#/c/188837/ [19:35:53] you can reset it [19:35:55] <^d> So Titan is 860%? [19:36:03] reset the addition of 'false &&' to logging.php, i mean [19:36:05] aha [19:36:08] since with https://gerrit.wikimedia.org/r/#/c/188837/ it's just false && false [19:37:10] thanks [19:43:10] 3MediaWiki-Core-Team: Set up mini wikifarm in Labs which has SecurePoll on it - https://phabricator.wikimedia.org/T88725#1018877 (10Deskana) 3NEW [19:44:19] csteipp: now an A, https://www.ssllabs.com/ssltest/analyze.html?d=creativecommons.org [19:44:41] 3MediaWiki-Core-Team: Set up mini wikifarm in Labs which has SecurePoll on it - https://phabricator.wikimedia.org/T88725#1018897 (10Deskana) [19:45:28] 3MediaWiki-General-or-Unknown, MediaWiki-Core-Team: Title::isValidMoveOperation() no longer respects $auth parameter - https://phabricator.wikimedia.org/T74870#1018900 (10Legoktm) [19:46:33] 3MediaWiki-Core-Team, MediaWiki-extensions-SecurePoll: Set up mini wikifarm in Labs which has SecurePoll on it - https://phabricator.wikimedia.org/T88725#1018919 (10bd808) Desired by 2015-02-13 or sooner. Needed to help gather requirements for possible SecurePoll changes for next major election. [19:46:44] 3MediaWiki-Core-Team, MediaWiki-extensions-SecurePoll: Set up mini wikifarm in Labs which has SecurePoll on it - https://phabricator.wikimedia.org/T88725#1018921 (10bd808) p:5Triage>3High [19:47:45] 3MediaWiki-Core-Team, MediaWiki-extensions-ConfirmEdit-(CAPTCHA-extension): Improve performance of Captcha extension on edit forms - https://phabricator.wikimedia.org/T88661#1018925 (10Legoktm) [19:48:43] 3MediaWiki-Core-Team: Hack on dumping infrastructure - https://phabricator.wikimedia.org/T88728#1018933 (10Manybubbles) 3NEW [20:02:19] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1019007 (10Fceller) >>! In T88549#1018451, @Fceller wrote: > >> There are thousands of indexable attributes, but not each document has every one of them - typical one would have tens,... [20:09:32] 3MediaWiki-Core-Team: Use parentRevId field for section change merging instead of timestamps - https://phabricator.wikimedia.org/T88734#1019036 (10aaron) 3NEW [20:19:00] 3MediaWiki-Core-Team, CirrusSearch: inefficient work of CirrusSearch in Russian Wikipedia - https://phabricator.wikimedia.org/T88724#1019060 (10bd808) [20:20:27] 3MediaWiki-Core-Team: Hack on dumping infrastructure - https://phabricator.wikimedia.org/T88728#1019064 (10bd808) This seems to have some similarities to the state of scap 12 months ago. Very heavily used home grown tech that is only understood by a small number of people. Seems like a great candidate for a "dev... [20:27:37] 3MediaWiki-General-or-Unknown, MediaWiki-Core-Team: Title::isValidMoveOperation() no longer respects $auth parameter - https://phabricator.wikimedia.org/T74870#1019078 (10Legoktm) 5Open>3Resolved [20:46:43] 3MediaWiki-Core-Team, CirrusSearch: inefficient work of CirrusSearch in Russian Wikipedia - https://phabricator.wikimedia.org/T88724#1019103 (10Manybubbles) Scratch that intitle: theory - it was wrong. [20:46:55] manybubbles: does elastic have any WAN support (e.g. write locally, replicate later)? [20:47:14] not really [20:47:26] if we wanted to do that the best way would be to rerun the cirrus jobs twice [20:47:34] yep, tee the jobs to both dcs [20:47:37] there are other options, probably, but they aren't as simple [20:47:46] with shared-nothing clusters [20:53:01] 3MediaWiki-Core-Team, wikidata-query-service: Investigate ArangoDB for Wikidata Query - https://phabricator.wikimedia.org/T88549#1019128 (10Smalyshev) @Fceller it's a partial import of JSON dump from http://dumps.wikimedia.org/other/wikidata/ (for some reason not all the data loaded, maybe due to some error in t... [21:06:46] manybubbles: does es have a concept of async-only replicas ? [21:07:07] well, basically slave nodes [21:30:38] 3MediaWiki-Core-Team: Hack on dumping infrastructure - https://phabricator.wikimedia.org/T88728#1019212 (10mark) Nice! This is exactly a conversation I wanted to start in a few weeks when I'm back. :) Yes, it would be really great if we could work with MW Core on improving our dumps infrastructure. There are lon... [21:39:02] 3MediaWiki-Core-Team: Use parentRevId field for section change merging instead of timestamps - https://phabricator.wikimedia.org/T88734#1019230 (10awight) woot! I've been doing a little work to prepare for this, I'll paste in my WIP patch and some suggestions for renaming internal variables to reduce confusion. [21:45:43] 3MediaWiki-Core-Team: Use parentRevId field for section change merging instead of timestamps - https://phabricator.wikimedia.org/T88734#1019276 (10aaron) [21:48:26] 3MediaWiki-Page-editing, MediaWiki-Core-Team: Use parentRevId field for section change merging instead of timestamps - https://phabricator.wikimedia.org/T88734#1019295 (10Ciencia_Al_Poder) [21:54:09] bd808: [Thu Feb 05 21:49:49.157740 2015] [:error] [pid 3785] [client 75.18.200.4:51277] PHP Notice: Array to string conversion in /srv/mediawiki/php-1.25wmf15/includes/debug/logger/legacy/Logger.php on line 298 [21:54:19] andrew is noticing that during this wikitech move to silver [21:54:24] should I file that as a task? [21:54:45] hmmm... what version of MW is that runnning? [21:56:07] honestly not sure [21:56:23] special version says wmf14 [21:58:35] though, that path uses wmf15 [21:58:50] I guess if he synced from tin, it's gonna be moved, whereas wikitech on virt1000 is likely behind [22:02:41] Reedy: looking at that line it looks like something is passing an array as the log message? [22:03:23] yeah [22:03:45] it's not really high priority issue [22:03:51] stack trace would be awesome if he can get one [22:04:18] ori: I can't find the right graph to verify that we aren't stupid for closing that perf issue on wmf15 bug [22:04:41] ori: that I'm not stupid for closing it that is, I'm about to since Reedy deployed the train today :) [22:04:47] the graphs died :( 75% not begin recorded now [22:04:58] * bd808 will see if he can make a new graph [22:05:21] * greg-g frowns [22:05:29] perfect storms everywhere [22:07:26] greg-g: https://graphite.wikimedia.org/render?from=-12hours&until=now&width=800&height=400&target=frontend.navtiming.mediaWikiLoadComplete.desktop.overall.mean [22:08:04] https://graphite.wikimedia.org/render?from=-1days&until=now&width=800&height=400&target=frontend.navtiming.mediaWikiLoadComplete.desktop.overall.mean [22:08:52] so deploy spike aside, the mean is hovering around 2s again. A bit up from a few days ago but not much. [22:08:57] https://graphite.wikimedia.org/render?from=-3days&until=now&width=800&height=400&target=frontend.navtiming.mediaWikiLoadComplete.desktop.overall.mean [23:32:27] ori: man, https://github.com/jonathanslenders/asyncio-redis looks cool [23:54:11] <^d> protip: `git clean -dfx` on a MW clone deletes lots of useful files [23:54:17] <^d> Like LocalSettings, vendor/ [23:55:45] use -i