[08:02] <Laney> vorlon: Good point, I'll add it to the agenda for the release sprint
[08:02] <Laney> had planned to add a GG-cycle discussion there too
[08:28] <mwhudson> jibel: thanks for reviewing my^WLaney's code :)
[08:29] <Laney> sil2100: ^- can you share the link to the doc you're prepping so I can add those items please? :-)
[08:29]  * Laney elbow bumps mwhudson 
[08:30] <Unit193> Hrm, is there somewhere I can push a package to see how Ubuntu's autopkgtest would handle it?
[08:31] <Laney> if you can upload: bileto, or you can trigger from a PPA: https://wiki.ubuntu.com/ProposedMigration/#Testing_against_a_PPA
[08:32] <jibel> mwhudson, np
[08:33] <Unit193> I'm only MOTU, so the second it is.  Thanks, Laney!
[08:34] <Laney> Unit193: Test using autopkgtest-buildvm-ubuntu-cloud locally first, though
[08:34] <Laney> For most cases that's close enough to what you'll get
[08:35] <Unit193> Laney: This shouldn't be one of those cases, I'm testing a specific quirk of the Ubuntu ADT stuff.  In theory, one should pass and one should gracefully skip.
[08:36] <Laney> I'm interested what quirks we have there
[12:56] <rbasak> cpaelzer: so on upgrade from Bionic to Focal, chrony is being removed in favour of systemd-timesyncd?
[13:17] <rbasak> cpaelzer: ^?
[13:17] <rbasak> Have I understood the bug correctly?
[13:18] <cpaelzer> hi rbasak
[13:19] <cpaelzer> rbasak: what you mention is bug 1872902
[13:19] <cpaelzer> which rbasak works on to fix in do-release-upgrader
[13:19] <cpaelzer> rbasak: if due to that (or any other condition) you have the new sytemd of focal and the old chrony
[13:19] <cpaelzer> rbasak: and then you remove the old chrony (which also happens on upgrade)
[13:19] <cpaelzer> then the crash triggers
[13:20] <rbasak> I think I follow, thanks.
[13:20] <cpaelzer> because the chrony postrm in bionic isn't tolerant to the case if systemd-timsyncd might be unsinatlled/masked/otherwise not there
[13:20] <rbasak> So this bug is still valid because if you remove chrony there's no longer any guarantee (and with the Conflicts it's impossible) that systemd-timesyncd will be available
[13:20] <cpaelzer> we added hardening a while ago but never needed it in bionic
[13:21] <rbasak> For 1872902, would a Conflicts/Replaces on chrony help?
[13:21] <cpaelzer> rbasak: for 1872902 no, that is what caused all of this  - it was properly split out of systemd to be a package on its own with proper conflicts/breaks and so on
[13:22] <cpaelzer> rbasak: the issue is the dependency resolution in d-r-u it seems
[13:22] <rbasak> "Replaces allows the packaging system to resolve which package should be removed when there is a conflict"
[13:22] <cpaelzer> a blunt bionic/focal upgrade without do-release-upgrade will not remove it
[13:23] <cpaelzer> rbasak: this is multi-dimensional with openntp/ntp/chrony/sytemd-timesyncd
[13:23] <rbasak> OK
[13:23] <cpaelzer> xnox:  and rbalint work on it and I'll leave it to them to pick their fix for 1872902 in focal
[13:23] <rbasak> Anyway, let me resolve the current SRU then
[13:24] <rbasak> Have I broken "git ubuntu clone chrony" in the edge channel?
[13:25] <cpaelzer> I only fetched in the recent days
[13:26] <cpaelzer> let me try
[13:26] <rbasak> Oh
[13:26] <rbasak> I think I might be running out of my source tree by accident
[13:26] <cpaelzer> it works for me on edge
[13:26] <cpaelzer> yep clone complete
[13:26] <rbasak> Thank you for checking
[13:27] <rbasak> Yep I've broken it in master I think, not published to edge yet
[13:30] <rbasak> Accepted, thanks!
[13:59] <Laney> bdmurray: any chance you could review https://code.launchpad.net/~marcustomlinson/update-manager/update-manager/+merge/382291 for marcustomlinson please? I got a bit close to writing the code for this one so I don't feel like a good reviewer any more. :-)
[14:31] <cpaelzer> mvo: stgraber: since it now re-occurred I guess I better ask you, is lxd being lockged (all lxd commands hang) until snapd was restarted a known issue
[14:31] <cpaelzer> I'm seeing this on s390x in 20.04
[14:32] <cpaelzer> I can't really reproduce, but it is the second time I found the system in that state
[14:32] <cpaelzer> I'd mostly want to know what to look out for next time I might hit the case
[14:33] <stgraber> Hmm, not something we've seen reported so far. Having to restart snapd is pretty suspicious though, maybe related to the apparmor profile ordering change that went into snapd recently?
[14:33] <cpaelzer> any logs in particular that you'd want?
[14:33] <cpaelzer> last time I saw it maybe 4 weeks ago, so I'd not look for a recent change
[14:34] <stgraber> 'ps fauxww', 'systemctl -a', 'snap changes' and the journal output would be a good start I think
[14:37] <bdmurray> Laney: sure, I can have a look at that
[14:41] <cpaelzer> stgraber: mvo: I filed bug 1873004 for now
[14:59] <mvo> cpaelzer: if you have the system in this state, does "SNAPD_DEBUG=1 lxc ..." show anything useful? can you attach the output of that as well?
[14:59] <mvo> cpaelzer: when you saw it last time, was it on s390x as well?
[15:05] <cpaelzer> mvo: yeah it always was on s390x so far
[15:05] <cpaelzer> I'll note the extra debug request on the bug for when it hangs agin
[15:05] <cpaelzer> again
[15:07] <hallyn> argh.  my ubuntu-core-dev membership expired :(
[15:07] <hallyn> the reminders were lost in spam
[15:07] <hallyn> do i have to go through the whole process again ? :(
[15:07] <hallyn> (i noticed bc people.ubuntu.com/~serge-hallyn seemed to disappear :( )
[15:08] <Laney> hallyn: email devel-permissions@lists.ubuntu.com and ask to be re-added
[15:09] <Laney> but if you haven't used your permissions in a while, maybe you want to ask for membership instead of upload access
[15:10] <hallyn> Laney: yeah, that's probably agood idea.  i wouldn't want to push anything without lots of review these days anyway
[15:10] <hallyn> thanks
[15:37] <gQuigs> xnox: did you see this? https://fedoraproject.org/wiki/Changes/systemd-resolved#DNSSEC   I read that as systemd giving up on DNSSEC..
[15:38] <xnox> gQuigs:  we have extra patches to make downgrade work better.
[15:38] <xnox> which i think are still not upstreamed
[15:38] <xnox> it is true the dnssec doesn't work still.
[15:39] <xnox> we do "maybe"
[15:39] <xnox> which i think is above DNSSEC=no, but below allow-downgrade.
[15:39] <xnox> not surprised
[15:40] <xnox> gQuigs:  i don't know if we need to change any of our defaults. But my guess is that dns-over-tls is taking over, because that works, and dns-sec does not.
[15:43] <gQuigs> rgr,I don't see as dns-over-tls doing the same thing, but I guess it's something.. thanks
[16:12] <cpaelzer> rbasak: FYI https://bugs.launchpad.net/ubuntu/+source/chrony/+bug/1872183/comments/12 in case you want to tweak the aging period anywhere that is the post to reply to
[16:19] <rbasak> cpaelzer: I commented. I think it's appropriate to release to bionic-updates now, if you're happy?
[16:19] <rbasak> cpaelzer: do we need the Focal fix to land first for this to be effective? I think not?
[16:20] <rbasak> ^ I'd appreciate a second opinion on skipping the ageing period for this SRU please, if someone's around.
[16:22] <cpaelzer> rbasak: the change has no dependency on anything else it has to wait on
[16:23] <popey> Is there some reason why we point directly to specific mirrors on a default install, and don't use mirrors://mirrors.ubuntu.com/mirrors.txt to load balance things? Is it something we have considered?
[16:25] <rbasak> cpaelzer: it does look like there are unrelated autopkgtest failure
[16:25] <rbasak> s
[16:26] <cpaelzer> rbasak: taking a look
[16:26] <cpaelzer> I was just working on the related test failures in focal :-) see bug 1873031
[16:27] <cpaelzer> looking at bionic now
[16:29] <cpaelzer> I know that bug, it seems someone backported some changes to linux headers into bionic
[16:29] <cpaelzer> rbasak: I can fix it but no more today
[16:29] <cpaelzer> rbasak: I'll provide a follow on upload to go over the current one once I had the time for it
[16:29] <cpaelzer> rbasak: I'll let you know tomorrow
[16:30] <rbasak> cpaelzer: sure, though is it necessary to fix this current bug?
[16:30] <cpaelzer> no it is unrelated
[16:30] <rbasak> I'm open to ignoring the test if appropriate
[16:30] <rbasak> What do you think?
[16:30] <cpaelzer> I think waiting a day is fine
[16:31] <rbasak> OK, I'll wait then
[16:31] <rbasak> Thanks!
[16:32] <cpaelzer> rbasak:  I added a comment to the bug so that anyone coming by knows
[16:51] <bdmurray> marcustomlinson: Can you explain a bit what happens in the cache.get_changes() loop?
[16:51] <bdmurray> re https://code.launchpad.net/~marcustomlinson/update-manager/update-manager/+merge/382291
[17:03] <bdmurray> or Laney ^^
[17:40] <marcustomlinson> bdmurray: we're looping through all packages that would be removed as a result of removing the replaced deb. If any of those have been manually installed we abandon the removal
[17:40] <marcustomlinson> e.g. in the case of snap-store replacing gnome-software. One may install ubuntu-software manually causing gnome-software to be marked auto
[17:41] <bdmurray> so ubuntu-software depended on gnome-software? I guess that's the part that wasn't clear to me.
[17:41] <marcustomlinson> bdmurray: yes
[17:41] <bdmurray> got it, okay I'll merge and upload
[17:41] <marcustomlinson> we'll run into more of this in the future. e.g. if we decide to replace libreoffice with the snap
[17:42] <marcustomlinson> we may look at libreoffice-core as the replaced deb
[17:42] <marcustomlinson> but one can install that via libreoffice, libreoffice-writer, etc.
[17:42] <marcustomlinson> thanks bdmurray
[18:15] <bdmurray> coreycb: If bug 1855890 is fixed can you update the status?
[18:39] <coreycb> bdmurray: done, thanks for the nudge
[20:31] <hallyn> Laney: hm, i suppose my ubuntu.com account was also tied to ubuntu-core-dev, so i won't get a response :)
[20:31] <hallyn> i'm so sad
[20:33] <hallyn> hm, no, that's not how this mail is set up.  oh well
[21:30] <CarlFK> wifi regression - hp laptop    Realtek  RTL8188EE cosmic is fine.  focal randomly (i guess) 261.278152] wlo1: deauthenticated and [ 2917.206570] rtlwifi: AP off, try to reconnect now
[21:31] <CarlFK> it just did that a bunch for a min or two, now is stable again.
[21:39] <CarlFK> what package should I log a bug?   dpkg -S rtl8188ee gives me a bunch of hits
[23:23] <rbasak> Where's the script that generates https://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses_by_team.yaml? It outputs a !!python/object/apply:collections.defaultdict which is a pain to parse safely. I'd like to switch it for a regular dict.
[23:44] <jbicha> infinity: mmm, I kinda want debhelper 13 in focal. That way we can build packages that have set debhelper-compat = 13