[02:13] <quadrathoch2> hey, is there a way to activate bash completions for snap containers? sitting on debian with lxd installed, and it's quite annoying without completions
[08:14] <pstolowski> morning
[08:17] <mvo> good morning pstolowski
[08:34] <pedronis> pstolowski: hi, is this another case of the bug we just fixed: https://bugs.launchpad.net/snapd/+bug/1919504 or is something else?
[08:34] <mup> Bug #1919504: panic: assignment to entry in nil map <snapd:New> <https://launchpad.net/bugs/1919504>
[08:39] <pstolowski> pedronis: looking
[08:45] <pstolowski> pedronis: so it panics in the same spot. it could be the same, unless there is some other edge case that triggers this, hard to say without more details. i'll ask in the bug report
[08:46] <pedronis> but the panic itself should be fixed?
[08:46] <pedronis> thanks
[08:48] <pstolowski> pedronis: yes
[09:00] <pstolowski> pedronis: when I was reproducing it last time snapd would eventually give up on the panic'ing tasks after a number of restarts; do you know what is actually making this happen in the taskrunner? I don't think we recover() (should we?)
[09:22] <pstolowski> mborzecki: hi, do you have time to take a look at #9959?
[09:22] <mup> PR #9959: o/snapstate: update validation sets assertions with auto-refresh <Needs Samuele review> <validation-sets :white_check_mark:> <Created by stolowski> <https://github.com/snapcore/snapd/pull/9959>
[09:57] <pedronis> pstolowski: Maciej is off today
[09:57] <pstolowski> ah, ok
[11:40] <mup> PR snapd#10046 opened: features: enable dbus-activation by default <Created by jhenstridge> <https://github.com/snapcore/snapd/pull/10046>
[11:45] <mup> PR snapd#10047 opened: tests: remove the snap profiler from the test suite <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/10047>
[12:00] <mup> PR snapd#10048 opened: [RFC] o/configstate: deal with no longer valid refresh.timer=managed <Needs Samuele review> <Created by stolowski> <https://github.com/snapcore/snapd/pull/10048>
[12:01] <pstolowski> mvo, pedronis ^ might be slightly controversial, let me know wdyt
[12:05] <mup> PR snapd#10049 opened: tests: remove snaps which are not used in any test <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/10049>
[12:17]  * pstolowski lunch
[12:45] <mup> PR snapd#10050 opened: tests: use snaps-state commands and remove them from the snaps helper <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/10050>
[12:50] <pedronis> pstolowski: I commented in it
[12:50] <pedronis> thx
[13:15] <pstolowski> pedronis: thanks, need to look into these other places
[13:19] <mvo> pstolowski: thanks for tackling this bug!
[13:22] <pstolowski> yw
[13:23] <jkridner_> ogra: I'm evaluating Ubuntu Core yet again, this time with a bit more interest as I finally have some experience using containers in a commercial project.
[13:24] <ogra> jkridner_, whee, awesome !
[13:25] <pstolowski> pedronis: btw have you seen my earlier question re taskrunner & panic/recover() ?
[13:26] <jkridner_> I'm trying to find the right docs on how to push images to boards once you boot them. The tutorial list is a bit overwhelming and not so step-by-step. And then, the first thing I encounter is you need HDMI for recovery.... not what I'd expect for an embedded system. Why not UART? https://ubuntu.com/tutorials/how-to-ubuntu-core-recovery-mode#1-overview
[13:30] <jkridner_> Can I run the PocketBeagle image on a BeagleBone Black (or Wireless, or Green Gateway)?
[13:35] <mup> PR snapd#10038 closed: tests: replace while commands with the retry tool <Created by sergiocazzolato> <Merged by sergiocazzolato> <https://github.com/snapcore/snapd/pull/10038>
[13:47] <pedronis> pstolowski: kind of, I'm not sure I understand it
[13:48] <pstolowski> pedronis: i'll try to explain on SU
[14:27] <ogra> jkridner_, i have never tried the pocketbeagle image on another board ... (it is set up with g_gether and g_serial so everything works through the USB cable, not sure the BBB can do that OOTB without adjustments, you'd have to try)
[14:28] <ogra> jkridner_, the image is pre-partitioned, to get it onto a board you'd just dd it to the target disk somehow (or for a start use an SD and simply dd it to that)
[14:29] <ogra> if you use UC20 it will re-partition during first boot ... UC16 and UC18 just expand the existing writable partition ...
[15:40] <pstolowski> mvo: i'm looking at https://github.com/snapcore/snapd/pull/10003 ; fwtw doing disable/mask/stop/unmask/enable/start manually works on 20.04; i wonder if there is a quirk with a specific systemd version
[15:40] <mup> PR #10003: configstate: add a config option for systemd-resolved <Needs Samuele review> <Created by woodrow-shen> <https://github.com/snapcore/snapd/pull/10003>
[15:49]  * cachio__ lunch
[15:56] <mvo> pstolowski: oh, interessting
[15:56] <mvo> pstolowski: let me quickly try on a uc20 machine
[15:57] <pstolowski> mvo: do you know what was the system they tried it on?
[15:58] <pstolowski> mvo: i tried both with --root=/ and without, worked both ways
[16:02] <mvo> pstolowski: I don't know where this did not work, I suspect spread?
[16:03] <mvo> pstolowski: oh, actually, let me look
[16:09] <mvo> pstolowski: I asked about more details in #20 on mattermost
[16:09] <mup> PR #20: Feature/snapfs cleanup kernel assets <Created by mvo5> <Closed by mvo5> <https://github.com/snapcore/snapd/pull/20>
[16:10] <mvo> pstolowski: also the unit tests in 10003 fail right now :/ this needs some more love it seems
[16:10] <pstolowski> mvo: yes i saw that
[16:11] <pstolowski> mvo: shall i take over this PR or are they on it?
[16:13] <mvo> pstolowski: I'm not on it, feel free to take it
[16:18] <pstolowski> mvo: ok
[16:26] <pstolowski> mvo: did you say there will be a meeting with them today?
[16:26] <pstolowski> mvo: i don't understand what is /run/systemd/resolve/stub-resolv.conf canary for in the test, it seems the PR is half-implemented
[16:27] <pstolowski> mvo: i can join if you want
[16:29] <mvo> pstolowski: I was talking with woodrow on mattermost, I don't think there will be a meeting today, it's also very late
[16:55] <ijohnson> mvo: pstolowski: should I give a quick shot at reproducing woodrow's bug?
[16:55] <ijohnson> it's very easy for me to test uc20 things :-D
[16:56] <pstolowski> ijohnson: i'd also be interested in knowing how to do that with kvm
[16:57] <pstolowski> ijohnson: if you could make a few notes documenting the steps i'd like to re-do it
[16:57] <ijohnson> pstolowski: sure, since in this case you are not resigning the kernel you don't need to rebuild anything there and can get on without tpm too, it's just rebuilding the snapd snap and use the documented instructions about kvm from https://docs.ubuntu.com/core/en/releases/uc20
[16:59] <ijohnson> note that you will probably have an issue trying to use console-conf if you just rebuild the snapd binary on something newer than focal and don't rebuild the entire snapd snap, since the implementation of getting users from extrausers is broken if you don't use cgo, but you can't use cgo on something newer than focal since then you are linking against a libc that's too new and it won't load
[17:00] <ijohnson> easy way to fix that is to just rebuild the whole snapd snap with lxd
[17:00] <ijohnson> but more time consuming obviously
[17:03] <pstolowski> ijohnson|lunch: i see, seems there will be a few hoops; and how do i sideload snapd?
[17:05] <ijohnson|lunch> hmm maybe I should put this into a doc, I need to get lunch
[17:06] <mvo> how is the store doing? is it worth re-running failing PRs?
[17:15] <pstolowski> ijohnson|lunch: yeah that would be great... please assume no prior knowledge of uc20 ☺️
[17:41] <mup> PR snapd#10005 closed: seed: ReadSystemEssentialAndBetterEarliestTime <Squash-merge> <Created by pedronis> <Merged by pedronis> <https://github.com/snapcore/snapd/pull/10005>
[17:56] <mup> PR snapd#9870 closed: vendor: update go-tpm2/secboot to latest version <Squash-merge> <Created by mvo5> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/9870>
[18:09]  * ijohnson may be intermittently available, internet service company tech here