16:00 #startmeeting 16:00 Meeting started Thu Mar 5 16:00:52 2015 UTC. The chair is slangasek. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 16:00 16:00 Available commands: action commands idea info link nick 16:00 [TOPIC] Lightning round 16:00 $ echo $(shuf -e barry doko stgraber jodh bdmurray slangasek caribou infinity mvo sil2100 robru cyphermox) 16:01 mvo robru bdmurray caribou slangasek barry cyphermox jodh infinity doko sil2100 stgraber 16:01 (base)vorlon@virgil:~/canonical$ 16:01 I win! 16:01 mvo: and go! 16:01 click: 16:01 - lp:~mvo/click/fix-autopkgtest and new 0.4.38.4 version 16:01 snappy: 16:01 - new snappy-go features: login (with 2fa), build 16:01 - no more dbus in snappy-go 16:01 - fixes, cleanup 16:01 - big refactor of ubuntu-core-upgrader, fix hardware.yaml unpack 16:01 misc: 16:01 - fix freeradius ftbfs and upstart reload bug 16:01 (done) 16:01 robru: 16:02 * CI Train 16:02 - refactoring continues, increased overall test coverage from 75 to 82% 16:02 - achieved 100% test coverage in merge_clean.py and branchhandling.py 16:02 - merge_clean.py shrank from 246 to 81 lines, without feature regressions. 16:02 - branchhandling.py shrank from 348 to 286 lines, also without feature regressions. 16:02 - fixed a bug affecting -gles packages where both gles and non-gles packages would clobber each other when being pushed to launchpad at publish time. 16:02 - fixed a bug preventing jobs from aborting when users abort jenkins jobs 16:02 - fixed bug in publish job which allowed it to silently succeed even though publication had failed. 16:02 - consolidated various version-mangling functions into shiny new version.py file 16:02 - fixed bug where build job was not timing out while waiting for packages to appear in ppa in certain conditions 16:02 - added 2 hour timeout on depwait packages, meaning there are no longer any conditions in which the build job can sit & spin forever. 16:02 - various and sundry cleanups, readability improvements, etc 16:02 * CI Engine 16:02 - slowly getting up to speed on the ticket system, submitted first branch towards replacing spreadsheet 16:02 barry: hi, the order is mvo robru bdmurray caribou slangasek barry cyphermox jodh infinity doko sil2100 stgraber 16:02 o/ 16:02 robru: it sounds like you have fun! 16:02 impressive results :) 16:02 mvo: thanks ;-) 16:03 robru: all done? 16:03 bdmurray: yeah 16:03 irc discussion with thedac regarding errors writing cores to swift 16:03 worked with thedac to increase --timeout for gunicorn daisy FEs to avoid OOPSes when writing cores to swift (set to 40s) 16:03 investigation into OOPSes when submitting cores to daisy 16:03 sorted out daisy / gunicorn logging issues 16:03 had daisy updated on staging and production to resolve logging issues 16:03 updated https://wiki.ubuntu.com/ErrorTracker/ServerArchitecture with information about more tables 16:03 updated daisy charm to properly set permissions for daisy app / gunicorn 16:03 investigation into how to discount package-from-proposed crashes in phased-updater regression check 16:03 worked on improving the fix for unattended-upgrades bug 1422345 per stgraber 16:03 bug 1422345 in unattended-upgrades (Debian) "stop being nice does not work" [Unknown,New] https://launchpad.net/bugs/1422345 16:03 submitted improved fix for unattended-upgrades bug to mvo on github 16:03 updated bug bot to better process SRU verification removal candidates 16:04 helped Riddell upload ubuntu-release-upgrader for vivid (test failure) 16:04 improved missing ddeb detector code to use a known list of missing ddebs and not notify about those 16:04 fixed error with linux apport hook and BootDmesg 16:04 bug management discussion with cyphermox 16:04 investigation into gvfs halted phased update 16:04 updated phased-updater code to include link to phasing report in email messages 16:04 sponsored patch to T,U fixing bug LP: #1422359 16:04 Launchpad bug 1422359 in heimdal (Ubuntu Utopic) "heimdal-kdc: kdc process leaks memory" [High,In progress] https://launchpad.net/bugs/1422359 16:04 bdmurray: thanks for the unattended-upgrades branch! 16:04 ✔ done 16:04 Bugfix : 16:04 mvo: is it merged? 16:04 - Misc SRU follow-up 16:04 bdmurray: *cough* not yet :( 16:04 - openstack : instances left dying : trying to reproduce 16:05 - openafs dkms build : SRU completed (LP: #1423151) 16:05 bdmurray: but soon! 16:05 (done)à 16:05 Launchpad bug 1423151 in openafs (Ubuntu Trusty) "openafs-modules-dkms 1.6.7-1: openafs kernel module failed to build" [Medium,Fix committed] https://launchpad.net/bugs/1423151 16:06 slangasek: all yours 16:06 * slangasek nods 16:06 * snappy alpha-3 milestone released this week 16:06 * reviewing 16:06 * partner calls regarding new server opportunities 16:06 * hiring: going through resumes now for Java engineer role, to schedule 16:07 interviews next week 16:07 * making progress on the phone channel refactoring proposal, will be 16:07 submitted for consideration this week 16:07 * systemd freeze exception: possibly flipping the switch today provided a 16:07 release FFe review 16:07 (done) 16:07 ok I have no idea what 'reviewing' in there was, ignore that line ;P 16:07 system-image: refined server merge proposal. LP: #1423622. si 3.0 betas to PPA 16:07 Launchpad bug 1423622 in Ubuntu system image "Please provide machine parsable output" [High,Fix committed] https://launchpad.net/bugs/1423622 16:07 qa sprint: ongoing 16:07 other: LP: #1363642 (debian bug #744145) 16:07 Launchpad bug 1363642 in python-pip (Ubuntu) "pip breaks after upgrading python package "requests"" [Undecided,Fix released] https://launchpad.net/bugs/1363642 16:07 Debian bug 744145 in python-pip "pip crashes on "import requests"" [Serious,Fixed] http://bugs.debian.org/744145 16:07 --done-- 16:07 * console-setup merge for 1.108: 16:07 - finish testing install, upgrade, etc. 16:07 - fixing issue with mangled keyboard/layout names 16:08 - fix typo in config script which broke layout selection skipping. 16:08 - FFe for upload 16:08 * Upload kbd to fix console-setup-mini being uninstallable. 16:08 * Removed console-setup-freebsd from console-setup (uninstallable) 16:08 * update kbd for console-setup-mini's new depends on console-setup-linux. 16:08 * debian-installer, installation-guide: 16:08 - update to use --- instead of -- as a separator 16:08 - this also fixes OEM mode, accessibility not working on server. 16:08 * update debian-cd to use --- as separator on built ISOs. 16:08 * discussed boot parameters for the installer with smoser. 16:08 - bug 1402042, and how it relates to -- vs. ---. 16:08 bug 1402042 in MAAS trunk "console= parameters need to be added before -- on kernel cmdline" [High,Triaged] https://launchpad.net/bugs/1402042 16:08 * debugging grub2 bug 1349538 16:08 bug 1349538 in Ubuntu "grub-install failure on 2T disk and msdos partition table" [High,Fix released] https://launchpad.net/bugs/1349538 16:08 * walked through/documented steps to test ubiquity changes with robert_ancell. 16:08 * reviewed some merges from Yu Ning for usb-creator. 16:08 * some discussion about BlueZ 5. 16:08 (done) 16:09 jodh: 16:09 * misc: 16:09 - Appraisal work. 16:09 * snappy: 16:09 - logging discussions. 16:09 - Promotion image testing. 16:09 - Raised MP for bug 1418966. 16:09 bug 1418966 in snappy-ubuntu "setting the hostname is not persistent accross reboots" [Undecided,In progress] https://launchpad.net/bugs/1418966 16:09 - Fixed bug 1428276. 16:09 bug 1428276 in snappy-ubuntu "wrong date formatting in 'snappy go list' output" [Undecided,In progress] https://launchpad.net/bugs/1428276 16:09 - Reviewed a bunch of branches and raised a few minor MPs. 16:09 - Finished adding more unit tests and functional tests to 16:09 ubuntu-core-upgrader (now landed). 16:09 - First pass at adding locking primitives to snappy 16:09 (including new test goodness :) 16:09 𐄍 16:10 infinity sent his regrets for this morning 16:10 doko: 16:10 - prepare binutils update for trusty-proposed 16:10 - gcc-4.9 update for vivid 16:10 - GCC 5: updates, reducing and filing upstream issues for ICEs 16:10 - golang, move gofmt to use alternatives 16:10 - handle a bunch of MIR's 16:10 - start fixing GCC 5 issues in packages 16:10 (done) 16:11 and sil2100 isn't in here 16:11 so stgraber 16:12 stgraber: here? 16:12 yep 16:12 [TOPIC] AOB 16:12 oh 16:12 [TOPIC] Lightning round 16:12 stgraber: 16:12 Short week, was off on Tuesday and Wednesday. 16:12 16:12 - LXD 16:12 - Released LXD 0.2 16:12 - Various feature work for 0.3 (reworked the exec code, fixed testsuite, translation, ...) 16:12 - Plenty of code reviews and bug handling 16:12 - Preparing for 0.3, due Tuesday 16:13 - System image 16:13 - Code review and discussion with Barry 16:13 - Admin paperwork (HR reviews, expense report, travel plans, ...) 16:13 16:13 (done) 16:13 thanks 16:13 [TOPIC] AOB 16:13 stgraber: did you look at the updated mp? 16:13 I have had a request for a MIR for openafs following a dkms breakage with the new Trusty kernel 16:13 barry: haven't yet, no 16:14 stgraber: k, no worries 16:14 the dkms build is not tested, hence a change in the dentry structure broke the openafs module rebuild 16:15 would such situation be better covered if openAFS would be included in main ? 16:15 caribou: how does the package being in main or not relate to the dkms build not being tested? 16:15 I think putting it in main is a rather heavyweight solution for addressing the package not being tested on kernel updates 16:15 slangasek: that's the question I have : having it in main would not give us better result on such cases 16:16 slangasek: especially since the src pkg build quite a chunk of packages. 16:16 because main implies security review and security support, when all it seems you're asking for is to keep the package from regressing 16:16 apw did sponsor the upload for the SRU so I did talk to him about this 16:16 slangasek: indeed. Good then, that's the answer I was after, thanks 16:16 caribou: so we now have autopkgtest support for SRUs 16:17 it would be good to have a dkms build as an autopkgtest in the openafs package (and I would accept an SRU for this) 16:17 slangasek: ok, I will look into this 16:17 then http://people.canonical.com/~ubuntu-archive/pending-sru.html would report the dkms failure ... er I think, unless the kernel packages are handled *too* specially on there 16:18 bdmurray: ^^ would we get revdep autopkgtest results displayed for the kernel? 16:18 oh, but the other problem is that openafs-modules-dkms doesn't actually depend on a kernel package, because that has side effects 16:18 slangasek: in that case, it would be the openafs-modules-dkms package that would fail, woudn't it ? 16:19 so step 1) openafs should have an autopkgtest to prove the dkms module is buildable 16:19 slangasek, i thought we had automatically implied autopkgtests for dkms things now 16:19 apw: oh? I don't know about this 16:19 what we don't have is implied relationships between the dkms packages and the kernel 16:19 caribou: yes, and the test failure of a reverse dependency is treated as a regression in the package that changed 16:20 slangasek: got it 16:20 apw: ok, so is that somewhere in the CI team's backlog? :) 16:21 caribou: anyway, hopefully that helps as far as identifying the "right answer"... there are probably still a few pieces to be put in place before it's fully implemented 16:21 o/ 16:21 slangasek, that i do not know, i know we do have dkms testing, and we have been talking about using the results there to block the kernle via the britney blocks 16:21 Sorry guys, my lunch shifted and it happened during the meeting 16:21 sil2100: hey there :) 16:21 I completely missed it... can I still put my status update? 16:21 ;) 16:22 apw: yes, exactly - though for SRUs we don't use britney so the blocks don't make a difference AFAIK 16:22 apw: on the other hand, you could mark the linux SRU bugs 'verification-failed'...? 16:22 apw: (with a bot) 16:22 (and a pointer to the test results) 16:22 sil2100: yes, go ahead 16:22 Ok, a quick one then from my side: 16:23 - Landing team work, silo coordination, preparing landing e-mails 16:23 - Documenting the early ww09-2015 milestone on the reports page 16:23 - Clean-up in blocker bugs and ubuntu-rtm silos 16:23 - Minor spreadsheet script fixes 16:23 - Send out request to developers regarding the QA sign-off procedures 16:23 * Announcements, annoucements, announcements 16:23 - Patch pilot sessions: 16:23 * Sponsored trusty upload for virtualbox, prepare bug to include SRU information 16:23 * Prepared a package for httptunnel, modified the patch, did a test-build in a PPA 16:23 * Looking into cryptkeeper proposed debdiff 16:23 - Ubuntu Engineering Live! 16:23 - RTM Status meetings 16:23 - Resolving a FTBFS for UITK -gles, offering packaging advice 16:23 - Prepare custom tarball for ubuntu-rtm mako (updated termminal) 16:23 (done) 16:23 Never again late lunch 16:23 sil2100: late lunch? or early dinner? 16:24 hm, depends, since our lunch is usually the main meal 16:24 interesting 16:24 So we usually eat something more like supper in the evening 16:25 sil2100: I didn't know there was a difference between dinner and supper ;-) 16:25 the main meal is dinner; the evening meal is supper; the noon meal is lunch 16:25 discuss 16:25 T. I. L. 16:25 or we could adjourn ;) 16:26 i prefer brinner 16:26 hah 16:26 brunch is good too :) 16:26 I've often enjoyed a good breakfunch ;-) 16:26 brlunner 16:26 but at all costs you should avoid lekfast 16:26 lol 16:26 3 meals in one 16:26 #endmeeting