## Template for LP Production Meeting logs. Just paste xchat log below and the format IRC line will take care of formatting correctly #format IRC #startmeeting Meeting started at 10:00. The chair is matsubara. Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] me :-D Welcome to this week's Launchpad Production Meeting. For the next 45 minutes or so, we'll be coordinating the resolution of specific Launchpad bugs and issues. [TOPIC] Roll Call New Topic: Roll Call me uh, me me rockstar, Chex, bigjools, allenap: hi me apologies from stub me me me [TOPIC] Agenda New Topic: Agenda matsubara: I'm sitting in for Chex this meeting as he's working on U1 stuff * Actions from last meeting * Oops report & Critical Bugs & Broken scripts * Operations report (mthaddon/Chex/spm/mbarnett) * DBA report (stub) * Proposed items thanks mthaddon [TOPIC] * Actions from last meeting New Topic: * Actions from last meeting * matsubara to file a bug on oops-tools to recognize new oops prefixes and sort out conflicting prefixes with losas * Chex to check app server logs and apache logs to see if it can shed any light in the high load issue. * adeuring to check with gmb about checkwatches failure * danilos to check bug 438039, assess if it's really critical. if it's is, land a fix, if it's not, update the importance * bigjools to investigate update-cache failure and reply back to the list Launchpad bug 438039 in rosetta "bzr branch import script oopses sometimes" [Critical,Fix released] https://launchpad.net/bugs/438039 matsubara: the bug tells you what it was :) oh, I forgot to 'me' myself I'll finish up my action today thanks danilos I chatted to curtis and as far as wecan tell it was caused by something else holding a transaction/table open not much I can do gmb replied to checkwatches failure email. it was a hung process which was killed and service resumed Since the PRF ran the following days, I believe it was a long running process that worried our watching proc bigjools, thanks for checking. I don't see new emails from that script failing so I take it's working normally yep mthaddon, any luck investigating the high loading issue? s/loading/load/ matsubara: I wasn't aware that was something we were following up on - not sure what the latest is, but I guess part of it plays into the new SplitIt stuff i.e. we've just brought a whole bunch of new servers online so we need to see what effect this has on the overall load of the system all right. I'll take that item off the list and if high load shows up in the graphs we can pursue further k thanks all, moving on [TOPIC] * Oops report & Critical Bugs & Broken scripts New Topic: * Oops report & Critical Bugs & Broken scripts it'sme gary_poster, bug 331990, can we CP it? * sinzui stares at Ursinha Launchpad bug 331990 in launchpad-foundations "The inline editor widget reports a JSON error when saving non-ASCII characters" [High,Fix committed] https://launchpad.net/bugs/331990 s/gary_poster/gary-sprint/ Ursinha: I do not have CP-foo. allenap, can we have a fix for bug 438802 and maybe CP it? gary-sprint, is this a matter of updating the lazr.restful lib used by lpnet? allenap, also, we have bug 438985, it's in progress but without activity for a some time Launchpad bug 438802 in malone "UnicodeDecodeError changing 'Assigned to' field when summary contains non-ascii" [High,Triaged] https://launchpad.net/bugs/438802 allenap, and bug 458180, that's BugTask index timeouts Launchpad bug 438985 in malone "Trying to make myself as bug supervisor of my project oopses" [High,In progress] https://launchpad.net/bugs/438985 Launchpad bug 458180 in malone "BugTask:+index timing out" [High,Triaged] https://launchpad.net/bugs/458180 sinzui, I've filed bug 458169 and bug 458189, the timeouts on Milestone and DistroSeries index pages rockstar, can we have a fix for bug 442981? Launchpad bug 458169 in launchpad-registry "Distroseries:+index page timing out" [High,Triaged] https://launchpad.net/bugs/458169 Launchpad bug 458189 in launchpad-registry "Milestone:+index pages timing out" [Undecided,New] https://launchpad.net/bugs/458189 Launchpad bug 442981 in launchpad-code "launchpad-project/+activereviews is OOPSing with TypeError (dup-of: 457541)" [High,Triaged] https://launchpad.net/bugs/442981 Launchpad bug 457541 in launchpad-code "Active code reviews for Loggerhead OOPSes on edge" [High,Fix released] https://launchpad.net/bugs/457541 Ursinha: maybe I misunderstood. are you asking for CP-blessing or for CP-shepherding? If the latter, sure, we can shepherd. gary-sprint, shepherding Ursinha: I replied that I beleive they are dups of 455812 brad is already working on it bug 455812 Launchpad bug 455812 in launchpad-registry "distroseries milestone timeout" [High,Triaged] https://launchpad.net/bugs/455812 hmmm matsubara: not sure, will ask leonardr. sinzui,I'll mark it as a dupe then, thanks not yet Ursinha, the fix for that bug is closing it as a duplicate of bug 457541 Launchpad bug 457541 in launchpad-code "Active code reviews for Loggerhead OOPSes on edge" [High,Fix released] https://launchpad.net/bugs/457541 oh [action] gary to talk to leonardr about cherry picking lazr.restful updates on lpnet for bug 331990 Ursinha, also, that bug is Fix Released. ACTION received: gary to talk to leonardr about cherry picking lazr.restful updates on lpnet for bug 331990 Launchpad bug 331990 in launchpad-foundations "The inline editor widget reports a JSON error when saving non-ASCII characters" [High,Fix committed] https://launchpad.net/bugs/331990 +1 rockstar,it still happens, how come? Ursinha, so yes, you may have it before it was asked. I have assign the distroseries +index to edwin. I think EdwinGrubbs and bac will find this is the same problem The oopses of the two new bugs look the the oopses I have been tracking in the older bug Ursinha, doesn't oops for me. rockstar, so the summaries are lying :) Ursinha, does this url oops for you? https://edge.launchpad.net/launchpad-project/+activereviews rockstar, well, it's loading... I'll keep my eye on it and if needed reopen it, right? rockstar, thanks allenap, hi :) Ursinha: I talk to deryck about getting bug 438802 fixed, and gmb about bug 438985. Launchpad bug 438802 in malone "UnicodeDecodeError changing 'Assigned to' field when summary contains non-ascii" [High,Triaged] https://launchpad.net/bugs/438802 Launchpad bug 438985 in malone "Trying to make myself as bug supervisor of my project oopses" [High,In progress] https://launchpad.net/bugs/438985 allenap, thanks Ursinha: Bug 458180 is a perennial problem. Launchpad bug 458180 in malone "BugTask:+index timing out" [High,Triaged] https://launchpad.net/bugs/458180 allenap, I see the main offender is bug #1 https://bugs.launchpad.net/ubuntu/+bug/1 (Timeout) yes, *sigh* Ursinha: Yeah, it always is :) Ursinha, yes, but I can't see how it'd get reopened. It was bad data, we fixededed the database records. Ursinha: and you just made it worse with a reference now :) danilos, yes, just to prove my point :P allenap, it still happens in other bugs too as per jono's email to launchpad-dev about ubuttu timing out allenap, there are some oopses not caused by #1 thanks allenap matsubara: Okay, as someone said, perhaps it's the +text interface. gary-sprint, the "buildbot failure in Launchpad on jscheck", is it severe? allenap, I briefly trawled the summaries and there are a some sofr timeouts on +text, but soft timeouts shouldn't be affecting ubottu matsubara, Ursinha: We need to do something more drastic to get the bug page quicker I think. Caching, etc, and that's coming alone. We've done a lot of the other things we can think of, but I'll discuss it with the team. matsubara: That's interesting. s/alone/along/ I see some emails from francis and rockstar talking about it, is there something that can or needs to be done? allenap, perhaps those timeouts are not being logged as OOPSes? similar to 500 we see eventually from apache gary-sprint, ^ to the 500 errors I mean Ursinha, gary-sprint, it is my belief that windmill sucks. Ursinha: it does not appear to be a problem in the basic buildbot setup at the moment. There are failures in the tests. This doesn't seem to be a foundations issue AFAICT. Björn may very well be able to help when he returns matsubara: Okay, I'm not sure what you mean, but we can talk about it after the meeting. right gary-sprint, thanks for the info allenap, sure thing. I'll find the bug I'm referring to matsubara: Thanks. [action] allenap and matsubara to talk about the timeouts on bug pages ACTION received: allenap and matsubara to talk about the timeouts on bug pages right, I'm done here rockstar: that's probably a given. The more interesting question is whether it sucks worse than the alternatives. My impression is no, but a champion could fight for an alternate view, . allenap, is it possible to ask for a cp for bug 438802 when it's fixed? Launchpad bug 438802 in malone "UnicodeDecodeError changing 'Assigned to' field when summary contains non-ascii" [High,Triaged] https://launchpad.net/bugs/438802 gary-sprint, sadly, there is no better alternative. Windmill sucks less than anything else out there. * salgado is now known as salgado-lunch :-) Ursinha: Sure. anmar was having problems yesterday with bugs with chinese chars, I think it's worth doing a CP thanks allenap Ursinha: np, thank you :) :) ok, two fix committed critical bugs rockstar, we had some failures on the update_preview_diffs script on the 19th matsubara, yeah, we're currently in the process of fixing the various oopses that script creates. * thumper has quit (Remote closed the connection) rockstar, ok. can you give me the bug numbers after the meeting? matsubara, there are many. gladly we have an oops tag to filter those :-) rockstar, I'll ping you after the meeting thanks everyone I think that's all for this section. thanks everyone [TOPIC] * Operations report (mthaddon/Chex/spm/mbarnett) New Topic: * Operations report (mthaddon/Chex/spm/mbarnett) SplitIt is the big one this week - now complete with exception of Auth DB split. New App servers brought online after haproxy throttlng of connections, we're watching how things are progressing A number of CPs done this week Is everyone clear on the new CP process? Shipit now managed by ISD, and CPs to be approved by nigelp Some app servers dying, loggerhead dying, poppy died once - is there a process for reviewing the Incident Log? That's about it mthaddon, last I heard Francis was the one to champion the Incident Log process. matsubara: basically we want to be sure someone's reviewing it to look for operational trends in production did he mention making trvial wiki edits for codebounce so we don't get email for those? ideally we won't need that codebounce all the time :-) matsubara, +1 :) bigjools: if we have to get alerts and go through the whole restart, edit wiki nightmare, you can put up with a few wiki edit notifications :) * matsubara looks at rockstar mthaddon: well, no I don't :) mthaddon: the concern is that we may learn to ignore it unless we can filter stuff out *we* can't do anything about mthaddon, I'm subscribed and get the pleasure of seeing every time you restart loggerhead. any news about the codebrowser dying all the time? what danilos said mthaddon: specifically, translations or soyuz team can't help much with codebrowse restarts matsubara, we are bringing someone on to look into the codebrowse issue. That's all I know. We certainly don't have the bandwidth currently to do it. fwiw, I usually do trivial that one - I guess maybe the other losas don't - will mention it bigjools: just as an example, and these are very, very common I believe there is a plan to but people to work on loggerhead mthaddon: in general, anything else shoudn't be a trivial edit, and codebrowse should, that would help old men like bigjools deal with their email :) ha sinzui: yeah, as flacoste mentioned today, I think we are having a contract that starts today or tomorrow that's great news mthaddon: but, do note that most team leads are subscribed to LPIncidentLog, and if one isn't, feel free to poke them about it danilos: k, thx that's all mthaddon ? yep all right. thanks everyone [TOPIC] * DBA report (stub) New Topic: * DBA report (stub) stub is on vacation and looks like the db is fine AFAICT so let's move on. [action] matsubara to talk to stub about the DBA report when he gets back ACTION received: matsubara to talk to stub about the DBA report when he gets back [TOPIC] * Proposed items New Topic: * Proposed items no new proposed items and I think that's all for today Thank you all for attending this week's Launchpad Production Meeting. See https://dev.launchpad.net/MeetingAgenda for the logs. #endmeeting