[15:00]  * slangasek waves
[15:00] <sil2100> o/
[15:00] <cyphermox> o/
[15:00] <rbalint> o/
[15:01] <slangasek> #startmeeting
[15:01] <meetingology> Meeting started Thu Sep  6 15:01:11 2018 UTC.  The chair is slangasek. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
[15:01] <meetingology> Available commands: action commands idea info link nick
[15:01] <rcj> o/
[15:01] <slangasek> [TOPIC] Lightning round
[15:02] <philroche> \o
[15:03] <slangasek> $ echo $(shuf -e slangasek bdmurray xnox tdaitx doko sil2100 rbalint infinity cyphermox mwhudson rcj philroche Odd_Bloke fginther juliank cshep)
[15:03] <slangasek> doko tdaitx fginther infinity philroche xnox slangasek cshep juliank Odd_Bloke bdmurray mwhudson sil2100 rbalint cyphermox rcj
[15:03] <slangasek> doko: are you at your computer?
[15:03] <philroche> slangasek: We need to add tobijk & aleks_bogdanov
[15:03] <slangasek> philroche: I was giving them a pass for their first meeting :)
[15:03] <philroche> Welcome tobijk & aleks_bogdanov :)
[15:04] <slangasek> tdaitx:
[15:04] <rcj> slangasek: and cshep
[15:06] <slangasek> fginther:
[15:06] <fginther> * Out one day for national holiday
[15:06] <fginther> * Finished work to improve the alignment of an image
[15:06] <fginther> * Finished work and merged cloud-init ssh config unit tests
[15:06] <fginther> * Revising a specification document related to partner image publication
[15:06] <fginther> * Cloud image build system vanguarding
[15:06] <fginther> ✔ done
[15:06] <fginther> infi<tab>
[15:06] <fginther> not here I guess
[15:06] <fginther> philroche:
[15:06] <philroche> * Onboarding new team members. Welcome chep, tobijk & aleks_bogdanov :)
[15:06] <philroche> * Cloud image modifications
[15:06] <philroche> (done)
[15:06] <philroche> xnox:
[15:07] <xnox> * continued work on zkey enablement
[15:07] <xnox> * s390-tools fix up #1790831
[15:07] <xnox> * tboot point release / fixes ftbfs / fixes stuck in proposed &
[15:07] <xnox>   migration to ssl1.1 #1731097
[15:07] <xnox> * kexec-tools migrated #1783086 as well as cryptsetup & systemd
[15:07] <xnox> ..
[15:07] <xnox> done
[15:08] <slangasek>  * short week, Labor Day
[15:08] <slangasek>  * SRUs (shim, grub, nvidia)
[15:08] <slangasek>  * proposed-migration:
[15:08] <slangasek>   * helped shepherd glibc through
[15:08] <slangasek>   * tracing various maintenance issues with autopkgtest runners
[15:08] <slangasek>   * misc autopkgtest fixes
[15:08] <slangasek>  * still working to get nvidia packages for clouds into their final form
[15:08] <slangasek> (done)
[15:09] <slangasek> cshep seems to be unavailable at the moment
[15:09] <slangasek> juliank:
[15:09] <juliank> * Fixed aptitude for boost1.67 transition (multi_index_container behavior change in 1.66)
[15:09] <juliank> * Fixed packagekit crashing in providesCodec (LP: #1790671)
[15:09] <juliank> * Fixed PEP8 regression in yesterday's update-manager blocking transition
[15:09] <juliank> * Started investigating the shim loading grub via iPXE in qemu (bug 1789319)
[15:09] <juliank> * Started working on manpage changes for networkctl & networkd-dispatcher to explain states (bug 1787495)
[15:09] <juliank> * Discussed how to make less packages manually installed
[15:09] <juliank> * Discussed apt/security/landscape interactions
[15:09] <juliank> (d0n3)
[15:10] <slangasek> Odd_Bloke:
[15:11] <slangasek> Odd_Bloke may be unavailable
[15:11] <slangasek> bdmurray:
[15:16] <xnox> hey
[15:16] <xnox> nothing at all
[15:16] <xnox> just bdmurray: and then me....
[15:16] <sil2100> No Brian?
[15:16] <sil2100> Should I go?
[15:17] <juliank> technical issues
[15:17] <xnox> sil2100, he is in the hangout
[15:17] <sil2100> Ah, hangout!
[15:17] <bdmurray> worked with webops regarding transition of daisy to mojo env
[15:17] <bdmurray> email to juliank re: u-r-u apt error string matching
[15:18] <bdmurray> investigation into why team to package mapping hadn't been updated
[15:18] <bdmurray> discussion with jdstrand re ubuntu-security-bugs team and pkg subscriptions
[15:18] <bdmurray> submitted RT re daisy-app server log rotation
[15:18] <bdmurray> submitted RT re rabbitmq server and graphite failing again
[15:18] <bdmurray> duplicate bug consolidation for LP: #1789879
[15:18] <bdmurray> testing of release upgrade failure bug reports
[15:18] <bdmurray> digging into getting instance count of u-u recursion error crash
[15:18] <bdmurray> reported appstream bug LP: #1790693 (discovered during upgrade testing)
[15:18] <bdmurray> investigation into size of retrace_amd64 queue (large due accepting 1/3 of crashes but not having 1/3 of retracers)
[15:18] <bdmurray>   also issues with delete'ing files from swift and therefore removing from the queue
[15:18] <bdmurray> reviewed and merged deej's nrpe changes to the daisy, retracer, and errors charms
[15:18] <bdmurray> modified logrotation jobs for daisy and errors app servers
[15:18] <bdmurray> started work on updates must be installed before release upgrade
[15:18] <bdmurray> ✔ done
[15:18] <bdmurray> sil2100: there now you can go
[15:18] <sil2100> - Sick for parts of Monday and Tuesday (shorter week)
[15:18] <sil2100> - Regular SRU reviews and releases
[15:18] <sil2100> - Fixed --extra-ppas for ubuntu-image classic builds
[15:18] <sil2100> - raspi3 u-i classic:
[15:18] <sil2100>   * Figuring out why suddenly the local builds stopped working
[15:18] <sil2100>     (I missed out some fixes when pushing to the PPA that I had locally before)
[15:18] <sil2100>   * Iterating on the image builds until there's a proper rootfs
[15:19] <sil2100>   * Work on fixing the gadget bits and boot-file creation (duh, the kernel is missing right now!)
[15:19] <sil2100> - core18:
[15:19] <sil2100>   * Some experimenting trying to build a local raspi3 core18 image for testing
[15:19] <sil2100>   * Fighting with core18 image builds due to breakages in snapd
[15:19] <sil2100>   * Minor PR reviews
[15:19] <sil2100>   * Adding rsync to the core18 snap for debugging
[15:19] <sil2100> - Some work on the britney SRU regression notification test-suite
[15:19] <sil2100> - Some canidate reviews
[15:19] <sil2100> (done)
[15:19] <sil2100> bdmurray: o/
[15:19] <xnox> sil2100, are ubuntu-image autopkgtest good? i saw them failing earlier....
[15:19] <rbalint> (short week)
[15:19] <rbalint> * fixed rax-nova-agent regression on Xenial
[15:19] <rbalint> * cosmic-proposed gardening
[15:19] <rbalint> * WSL updates
[15:19] <rbalint> * working on LP: #1783023 but no fix yet
[15:19] <rbalint> (done)
[15:20] <rcj> cyphermox:
[15:20] <cyphermox> - shim-review: DriveLock
[15:20] <cyphermox> - shim 15 update and SRUs
[15:20] <cyphermox> - preparing netplan.io 0.40 release
[15:20] <cyphermox> - optional-addresses upstream systemd work
[15:20] <cyphermox> - gnu-efi SRU everywhere to support new shim, unbreak shim 13 in trusty.
[15:20] <cyphermox> - test rebuilds for gnu-efi revdeps
[15:20] <cyphermox> - grub2: sponsor add_ext_lfb_base_support.patch: Add support for ext_lfb_base. (LP: #1785033)
[15:20] <cyphermox> - investigating how to retrieving keys to validate kernel signatures
[15:20] <cyphermox> (done)
[15:20] <rcj> rcj:
[15:20] <rcj> * on-boarding a new team-mate (Welcome cshep!)
[15:20] <rcj> * out on Monday for federal holiday (alas not a "bank holiday" philroche)
[15:20] <rcj> * meeting with cloud partners
[15:20] <rcj> ☑
[15:24] <slangasek> any questions on status?
[15:25] <slangasek> [TOPIC] Bugs
[15:25] <slangasek> bdmurray:
[15:25] <bdmurray> So there's this upgrade bug which is somehwat vague
[15:25] <bdmurray> https://bugs.launchpad.net/ubuntu/+source/ubuntu-release-upgrader/+bug/1789977
[15:26] <bdmurray> There was an lvm2-activation segfault during or after the upgrade
[15:26] <xnox> ouch
[15:26] <bdmurray> I managed to find a crash about lvm2 in LP and a bucket in the error tracker.
[15:27] <bdmurray> I'm not positive they are all the same issue though.
[15:28] <juliank> I don't see an actual upgrade bug in the log, do you?
[15:28] <juliank> regarding the dpkg --configure -a
[15:28] <bdmurray> Given the quantity of crashes about lvm2 I think that should be looked at by xnox though.
[15:28] <bdmurray> ;-)
[15:29] <juliank> I can say it's not crashing for me :)
[15:29] <bdmurray> juliank: maybe the screenlog would help?
[15:30] <juliank> I was looking at AptTermLog to see which packages remained unconfigured
[15:31] <slangasek> ok I think we'll take that one into trello
[15:31] <slangasek> bdmurray: next?
[15:32] <bdmurray> Bug 1767918 is a quick fix, just need to schedule it
[15:32] <slangasek> +1 I'll pull it into backlog now
[15:33] <slangasek> bdmurray: next?
[15:33] <bdmurray> Maybe the desktop team should look at the comments in bug 1788220?
[15:35] <rbalint> can we as Foundations do anything regarding shipping or not shipping things as snaps?
[15:35] <xnox> rbalint, on cloud images yes
[15:38] <slangasek> well, it's not something that's going to be reverted wrt 18.04
[15:38] <slangasek> if there were suggestions for concrete improvement to the implementation, that'd be one thing
[15:38] <bdmurray> I was surprised by files in $HOME/snap
[15:39] <xnox> the $HOME/snap is akward
[15:39] <slangasek> it is surprising if you were expecting ~/. but there are arguments and that's not a decision that's going to be reverted via a bug on gnome-calculator in launchpad
[15:39] <xnox> i would expect that on the server, but on desktop i would expect liek ~/.config/snap or ~/.local/snap
[15:39] <slangasek> bdmurray: +1 for asking the desktop team to field this, which is ultimately 'opinion'
[15:39] <xnox> because ~/.local/snap is ~+ /snap
[15:39] <bdmurray> slangasek: ack
[15:39] <xnox> because ~/.local/snap is ~= /snap
[15:40] <slangasek> (also I think that 1.6GB disk figure is incorrect)
[15:40] <bdmurray> bug 1754294 seems to be affecting some people when upgrading, while they are not ubuntu packages I wonder if there is anything we can do better there.
[15:41] <bdmurray> slangasek: That disk estimate for upgrading would need to take into account the space in /home too
[15:42] <slangasek> bdmurray: sorry, is that last comment referring still to the snap thing?
[15:42] <slangasek> because snaps only use space in /home after they've been executed by the user, so not an upgrade thing
[15:42] <bdmurray> slangasek: yes, ah okay
[15:42] <cyphermox> 1.6Gb figure does sound rather large
[15:43] <cyphermox> larger even than if you also had extra bits, classic snap, etc.
[15:43] <xnox> cyphermox, i think the killer is three copies mounted by default.
[15:43] <slangasek> cyphermox: I did the math and get 262M for Ubuntu Desktop 18.04.  I think the user is counting disk space under /snap improperly
[15:44] <cyphermox> heh
[15:44] <xnox> which i think makes sense on iot, but doesn't make musch sense on a desktop with always-on interent.
[15:44] <xnox> ogra, why is 3 copies, instead of just 2 for the desktop snapd deploys?
[15:46] <slangasek> bdmurray: the user is complaining that we are not shipping an ABI-compatible version of the library; the library ABI has changed because time moves forward
[15:46] <slangasek> bdmurray: we are not going to provide the old ABI, so those packages are going to be removed on upgrade, unless it makes the upgrade itself blow up there's nothing to do
[15:49] <bdmurray> slangasek: ack, that's it from me then
[15:49] <slangasek> ok
[15:49] <slangasek> [TOPIC] AOB
[15:49] <slangasek> anything else?
[15:50] <xnox> slangasek, i believe it did blow up upgrade
[15:50] <xnox> slangasek, cause php and something rather tried to load both inside mod_php and than apache would fail
[15:50] <cyphermox> no aob from me
[15:50] <xnox> slangasek, because some niche stack was not ported, and it should be reverted to the old abi
[15:50] <xnox> done
[15:51] <rcj> no aob from me
[15:51] <slangasek> #endmeeting
[15:51] <meetingology> Meeting ended Thu Sep  6 15:51:24 2018 UTC.
[15:51] <meetingology> Minutes:        http://ubottu.com/meetingology/logs/ubuntu-meeting/2018/ubuntu-meeting.2018-09-06-15.01.moin.txt
[15:51] <slangasek> thanks, all
[16:07] <xnox> slangasek, ie. i believe https://bugs.launchpad.net/ubuntu/+source/xmltooling/+bug/1776489 is valid
[16:07] <xnox> slangasek, and we should revert that stack back to libcurl3 which we do have available in bionic
[16:07] <xnox> slangasek, cause it can't apache php something rather
[16:08] <xnox> or like make libcurl3 co-installable somehow
[16:11] <slangasek> xnox: "that stack" meaning what, all of apache?
[16:12] <xnox> slangasek, no, not all of apache, "shibboleth" packages
[16:13] <xnox> i believe there is exec boundary, thus it's all fine, as long as one preloads or have the old-abi libcurl3 non-conflicting, or like recompile these 5 packages against the old one.
[16:13] <xnox> cause in practice it is broken with the new one, as it needs ssl1.0
[16:13] <xnox> behind libcurl api
[16:14] <slangasek> xnox: I thought the shibboleth stack /was/ left on libcurl3
[16:16] <xnox> yes, but not recompiled against the new libcurl3-openssl-reintroduced hence stuck with old-deps which are conflicting with curl4 and thus not co-installable with like any ubuntu install
[16:17] <xnox> or possibly our libcurl3-openssl-reintroduced is itself too-conflictly, than should have been.
[16:17] <slangasek> xnox: it's conflicty because it's the same soname (as per upstream) but different ABI
[16:18] <slangasek> so yeah
[16:18] <xnox> so we need to introduce that old abi under a new name, to make this stack work, imho.
[16:19] <xnox> rpath for the win!
[20:05] <pleia2> anyone here for the membership meeting?
[20:05] <Wild_Man> Here
[20:11] <tsimonq2> Sort of, ish.
[20:12] <Wild_Man> Is the membership applicant here?
[20:17] <popey> *tumbleweed*