| === jamesh__ is now known as jamesh | ||
| mup | PR snapcraft#3004 opened: storeapi: add channel-map endpoint <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/3004> | 01:22 |
|---|---|---|
| mup | PR snapcraft#3003 closed: repo: always use host source lists and remove those found in plugins <Created by cjp256> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3003> | 01:46 |
| mup | PR snapcraft#3001 closed: repo: type annotations and mypy fixes for base <Created by cjp256> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3001> | 03:37 |
| mup | PR snapcraft#3002 closed: repo: use functools.lru_cache for dpkg -L queries <Created by cjp256> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3002> | 03:37 |
| mborzecki | morning | 05:08 |
| mborzecki | finally some green in PRs, however the status from travis was not updated for somer PRs even though the actual job was successful | 05:34 |
| mborzecki | mvo: hey | 06:07 |
| mborzecki | mvo: the LXD PR is green | 06:07 |
| mvo | mborzecki: nice | 06:12 |
| mvo | mborzecki: hm, that's strange, I don't see an update | 06:12 |
| * mvo is confused | 06:12 | |
| mborzecki | mvo: hm? try refreshing the page | 06:26 |
| mup | PR snapcraft#3005 opened: Change dotnet runtime version from snapcraft.yaml <Created by amka> <https://github.com/snapcore/snapcraft/pull/3005> | 06:41 |
| zyga | good morning | 06:47 |
| zyga | man it's late today | 06:47 |
| mborzecki | zyga: hey | 06:47 |
| zyga | anything eventful | 06:49 |
| zyga | I EOD and then worked more and EOD after midnight | 06:49 |
| zyga | how are you mvo? did you sleep at all? | 06:50 |
| * zyga sips coffee and wakes up | 06:54 | |
| zyga | -3 outside, no people | 06:54 |
| mvo | zyga: hey, I'm good, thank you | 06:58 |
| pstolowski | morning | 07:01 |
| zyga | hey pawel | 07:06 |
| zyga | pstolowski: do you walk your dog in the morning? | 07:06 |
| zyga | was it all snowy as well? | 07:06 |
| pstolowski | zyga: no, it's cold but sunny | 07:07 |
| zyga | it was white outside here | 07:08 |
| zyga | so weird | 07:08 |
| zyga | mborzecki: hmmm, command foo should run foo without using built-in things, right? | 07:18 |
| mborzecki | zyga: built in? | 07:18 |
| zyga | built into shell | 07:19 |
| zyga | echo vs /usr/bin/echo or /bin/echo or whatever | 07:19 |
| zyga | command %q foo works in bash | 07:19 |
| zyga | er | 07:19 |
| zyga | command printf %q foo works in bash | 07:19 |
| zyga | /usr/bin/printf %q foo works in anything | 07:19 |
| zyga | but command printf %q foo doesn't work in dash | 07:19 |
| * zyga re-reads the man page | 07:19 | |
| mborzecki | oh dash, idk ;) | 07:20 |
| zyga | it's going for the builtin anyway | 07:20 |
| mborzecki | zyga: there's probably some footnote in posix spec that say why it should do that | 07:20 |
| zyga | ah sorry | 07:21 |
| zyga | command is not what I think it is | 07:21 |
| zyga | it just skips functions | 07:21 |
| zyga | brb | 07:22 |
| zyga | it passes :DDDD | 07:41 |
| zyga | I need to watch over Lucy for two hours | 08:11 |
| zyga | I'll be back later | 08:11 |
| zyga | sorry | 08:11 |
| mup | PR snapd#8385 opened: tests/main/user-session-env: stop the user session before deleting the test-zsh user <Created by jhenstridge> <https://github.com/snapcore/snapd/pull/8385> | 08:13 |
| jamesh | mborzecki: ^^^ I wouldn't mind a review of the above, since it is for a test you wrote | 08:14 |
| mborzecki | jamesh: sure | 08:14 |
| zyga | jamesh: reviewed | 08:18 |
| jamesh | zyga: thanks | 08:24 |
| mborzecki | heh, so i added the changes to do an immediate system reboot when install is complete, and the whole process feels much faster now :P | 08:27 |
| zyga | mborzecki: nothing like reboot to speed up computers ;) | 08:32 |
| * zyga has IRC but needs to pay attention to lucy until wife is back | 08:33 | |
| mup | PR snapd#8386 opened: many: support immediate reboot <UC20> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/8386> | 08:34 |
| mborzecki | rebooting to recovery feels nice too now | 08:35 |
| mborzecki | it will also shave some time off the uc20 test prepare phase | 08:36 |
| mborzecki | ehhh Mar 31 07:05:20 mar310653-786350 50-motd-news[3511]: /etc/update-motd.d/50-motd-news: 131: /etc/update-motd.d/50-motd-news: cannot create /var/cache/motd-news: Read-only file system | 08:36 |
| mborzecki | that's why google:ubuntu-core-16-64:tests/main/degraded is failing | 08:37 |
| mborzecki | why do we even have motd-news.service there? | 08:37 |
| mborzecki | mvo: ^^ do you recall why it's included? | 08:37 |
| mborzecki | oh and `HTTP/1.0 503 Service Unavailable` when POSTing to https://api.snapcraft.io/v2/snaps/refresh | 08:38 |
| jamesh | pedronis: I think I've covered everything in https://github.com/snapcore/snapd/pull/5822 (user-daemons PR). The intermittent failure was due to an existing test not properly cleaning up. | 08:38 |
| mup | PR #5822: wrappers: allow user mode systemd daemons <:birthday:> <Needs Samuele review> <⛔ Blocked> <Created by jhenstridge> <https://github.com/snapcore/snapd/pull/5822> | 08:39 |
| mvo | mborzecki: in a meeting right now | 08:51 |
| zyga | re | 09:02 |
| pedronis | jamesh: thansk, I will review it again tomorrow | 09:03 |
| pedronis | *thanks | 09:03 |
| jamesh | pedronis: thanks | 09:03 |
| zyga | mborzecki: same bug we had in core18 | 09:03 |
| mborzecki | hmmm, let me see the fix there | 09:03 |
| mup | PR snapd#8387 opened: store: search v2 tweaks <Simple 😃> <Created by stolowski> <https://github.com/snapcore/snapd/pull/8387> | 09:13 |
| zyga | eh, fun | 09:20 |
| zyga | core systems don't have writable /var/lib/systemd | 09:20 |
| zyga | so they cannot get /var/lib/systemd/linger | 09:20 |
| mvo | mborzecki: re motd-news.service> no idea, sounds wrong | 09:21 |
| zyga | mvo: same bug we had in core16, core18 and now core20 | 09:21 |
| mvo | zyga: uh, what is linger ? | 09:21 |
| zyga | mvo: it's something we just inherit from ubuntu | 09:21 |
| mborzecki | mvo: it's removed in core20 and core18, but not core | 09:21 |
| mvo | mborzecki: aha, sounds like an oversight then | 09:21 |
| zyga | mvo: linger is a directory used by systemd to indicate that a user may have session services running while not being logged in | 09:21 |
| mvo | zyga: I see, we can add it to the writable dirs if it's needed | 09:22 |
| zyga | mvo: I need it to make test non-racy | 09:22 |
| zyga | mvo: I think I can cope with test specific workaround but let's keep that option open | 09:22 |
| mborzecki | actually the fix in core18 is weird, https://github.com/snapcore/core18/blob/master/hooks/014-set-motd.chroot#L27-L31 | 09:22 |
| mborzecki | it removes the services and then update is disabled? | 09:22 |
| mborzecki | fwiw, the timers.target.wants/motd-news.timer symlink is left behind in 18 and 20 | 09:22 |
| zyga | mborzecki: nice | 09:23 |
| zyga | let's fix it :) | 09:23 |
| zyga | I think the services should have conditions | 09:23 |
| zyga | instead of hooks remving them | 09:23 |
| zyga | it would "scale" better | 09:23 |
| zyga | and since it's a ubuntu thing IIRC, we should have no problems adding that | 09:23 |
| mborzecki | duh | 09:25 |
| mborzecki | so update-motd is hooked up to pam? wtf? | 09:25 |
| zyga | whar? | 09:28 |
| zyga | how? | 09:28 |
| mborzecki | zyga: see pam_motd | 09:28 |
| zyga | and we are struggling with PATH | 09:29 |
| zyga | we should just add pam_snapd | 09:29 |
| zyga | ;D | 09:29 |
| zyga | more lucy time | 09:29 |
| mup | PR core#111 opened: live-build/hooks/motd: disable dynamic motd, disable motd services <Created by bboozzoo> <https://github.com/snapcore/core/pull/111> | 09:35 |
| mup | PR core18#149 opened: hooks/motd: cleanup dangling symlink, fix typo <Created by bboozzoo> <https://github.com/snapcore/core18/pull/149> | 09:37 |
| mup | PR core20#31 opened: hooks/motd: disable dynamic motd, cleanup dangling symlink <Created by bboozzoo> <https://github.com/snapcore/core20/pull/31> | 09:40 |
| * zyga spawns a test for core workaround and goes back to luck | 09:41 | |
| mup | PR snapd#8388 opened: cmd/snap,seed: validate full seeds (UC 16/18) <Created by pedronis> <https://github.com/snapcore/snapd/pull/8388> | 09:44 |
| pedronis | mvo: ^ | 09:45 |
| mvo | pedronis: thanks | 09:46 |
| zyga | ok, core systems now pass | 09:54 |
| zyga | only centos-7 and amazonlinux with ancient printf | 09:54 |
| zyga | let's reimplement %q | 09:54 |
| popey | My laptop locked up last night, had to hard-reboot. Now no snaps will run | 09:55 |
| popey | cannot change profile for the next exec call: No such file or directory | 09:55 |
| popey | snap-update-ns failed with code 1: No such file or directory | 09:55 |
| popey | Anyone seen that before? | 09:55 |
| zyga | popey: I would help but check my Twitter | 09:56 |
| popey | i would check your twitter but I can't open my browser | 09:56 |
| zyga | Maybe nothing mounted | 09:56 |
| zyga | Maybe partial dpkg update? | 09:57 |
| zyga | can you get root | 09:58 |
| popey | i can | 09:58 |
| popey | no dpkg in flight, i see squashfs files mounted | 09:58 |
| zyga | Can you dpkg —configure -a | 09:58 |
| zyga | Can you snap list? | 09:58 |
| popey | noting | 09:58 |
| popey | *nothing | 09:58 |
| popey | https://paste.ubuntu.com/p/MPDMzTvDNy/ | 09:59 |
| zyga | Snaps not broken ok | 09:59 |
| zyga | Can you snap changes? | 09:59 |
| zyga | Anything in flight? | 09:59 |
| popey | https://paste.ubuntu.com/p/JNkKSHBGqn/ | 09:59 |
| popey | nope | 09:59 |
| zyga | Can you snap run something? | 10:00 |
| popey | cannot change profile for the next exec call: No such file or directory | 10:00 |
| popey | snap-update-ns failed with code 1: File exists | 10:00 |
| zyga | Can you systemctl status apparmor.service | 10:01 |
| pedronis | pstolowski: some comments to #8387 | 10:01 |
| mup | PR #8387: store: search v2 tweaks <Simple 😃> <Created by stolowski> <https://github.com/snapcore/snapd/pull/8387> | 10:01 |
| popey | https://paste.ubuntu.com/p/mTqRsQJMKY/ | 10:02 |
| pstolowski | pedronis: ty | 10:02 |
| zyga | Can you as-status please? | 10:02 |
| zyga | (Tying form memory, hope that is the one) | 10:02 |
| popey | i dont think that's it | 10:04 |
| mborzecki | hmm snapcraft segfault? | 10:04 |
| zyga | mborzecki: help, afk, help popey list loaded aa profiles please | 10:05 |
| mborzecki | popey: sudo aa-status | 10:05 |
| popey | https://paste.ubuntu.com/p/7WD8FCFdRJ/ | 10:05 |
| zyga | popey: does the revision of core in that output match what the current points to on your system? | 10:06 |
| zyga | And is the list of snaps the same as those you see in that output? | 10:07 |
| popey | yes | 10:07 |
| popey | that's not a long list of snaps | 10:07 |
| zyga | Ok | 10:07 |
| popey | snap list shows me 88 snaps | 10:08 |
| zyga | Can you SNAP_CONFINE_DEBUG=yes snap run something please | 10:08 |
| mup | PR snapd#8332 closed: gadget: SystemDefaults helper for FilesystemOnlyApply (2/N) <Created by stolowski> <Merged by stolowski> <https://github.com/snapcore/snapd/pull/8332> | 10:09 |
| popey | https://paste.ubuntu.com/p/yJjr8NRzZM/ | 10:09 |
| zyga | Inspecting... | 10:10 |
| zyga | So, snap-update-ns.null is absent | 10:11 |
| zyga | Can you look at /var/lib/snapd/apparmor/profiles | 10:11 |
| zyga | Is it there? | 10:12 |
| popey | yes, lots in there | 10:12 |
| zyga | And that list is larger than aa-status, right? | 10:12 |
| popey | yes | 10:12 |
| zyga | Can you look at journal log of apparmor.service | 10:13 |
| zyga | I wonder if apparmor loaded stuff | 10:13 |
| zyga | At the same time snapd rewrote profiles | 10:14 |
| zyga | To fix tour machine please restart apparmor.service | 10:14 |
| * popey tries that | 10:14 | |
| zyga | Sorry for the silly typos from phone correction | 10:14 |
| popey | aa-status is much longer now | 10:14 |
| zyga | Check if that makes null work | 10:15 |
| popey | i can run snaps | 10:15 |
| zyga | Look through timeline | 10:15 |
| popey | give your baby a hug from me | 10:15 |
| zyga | Of startup uf | 10:15 |
| zyga | Apparmor servir | 10:15 |
| zyga | Apparmor service | 10:15 |
| zyga | And snapd service | 10:15 |
| zyga | Maybe we are racing? | 10:15 |
| zyga | If they overlap that is a bug | 10:15 |
| zyga | And thank you for reporting | 10:15 |
| popey | :D | 10:15 |
| popey | a bug in snapd or apparmor? | 10:16 |
| zyga | She is hugging me constantly now (sleeping) | 10:16 |
| zyga | In our units | 10:16 |
| zyga | We may need after= | 10:16 |
| zyga | So please check the time stamps carefully | 10:16 |
| zyga | Also | 10:16 |
| pedronis | mborzecki: I started reviewing your immediate reboot PR but need to go have lunch first | 10:16 |
| zyga | Time stamps on those files in /v/l/a/profiles | 10:16 |
| mborzecki | pedronis: ok, enjoy your lunch :) | 10:17 |
| * pstolowski lunch & errand | 10:19 | |
| mborzecki | ehh the core build process :/ | 10:21 |
| zyga | it passes on centos | 10:32 |
| zyga | running final check on amazon | 10:33 |
| zyga | liberated :) | 10:43 |
| zyga | we should fix 20 images | 10:48 |
| zyga | GDM is running | 10:48 |
| zyga | I bet that's not free | 10:48 |
| zyga | mvo: ^ | 10:48 |
| zyga | ubuntu-20.04-64 has GDM running | 10:48 |
| mvo | zyga: uh, nice find | 10:52 |
| mvo | zyga: let's talk to sergio about this | 10:52 |
| zyga | we know about this | 10:58 |
| zyga | just reminding because it affects a test | 10:58 |
| zyga | it's pulled in by evolution-data-server | 10:58 |
| zyga | mvo: I'll talk to him today | 11:01 |
| pedronis | mborzecki: I did a pass | 11:20 |
| mborzecki | pedronis: thanks, let me check the comments | 11:20 |
| pedronis | mborzecki: it looks good, mostly comments on details | 11:20 |
| ijohnson | hey folks | 11:47 |
| pstolowski | hey ijohnson | 11:50 |
| zyga | hey ian :) | 11:51 |
| ijohnson | hey zyga pstolowski | 11:52 |
| ijohnson | how's it going ? | 11:52 |
| zyga | good | 11:53 |
| zyga | making progress today | 11:53 |
| ijohnson | nice, my internet kept cutting out for a couple minutes at a time yesterday, so I'm hoping today it's more stable | 11:55 |
| mborzecki | pstolowski: please take another look at #8376 | 12:04 |
| mup | PR #8376: daemon: make POST /v2/systems/<label> root only <Simple 😃> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/8376> | 12:04 |
| pstolowski | sure | 12:04 |
| mborzecki | thanks! | 12:04 |
| mup | PR snapd#8389 opened: tests: make session tool way more robust <Test Robustness> <Created by zyga> <https://github.com/snapcore/snapd/pull/8389> | 12:04 |
| pstolowski | +1 | 12:07 |
| jdstrand | e/win 5 | 12:14 |
| pedronis | pstolowski: I reviewed your PRs | 12:39 |
| pstolowski | pedronis: thank you | 12:40 |
| pstolowski | mborzecki: perhaps you could make 2nd review of #8343? | 12:43 |
| mup | PR #8343: config, features: move and rename config.GetFeatureFlag helper to features.Flag <Created by stolowski> <https://github.com/snapcore/snapd/pull/8343> | 12:43 |
| mborzecki | pstolowski: sure, will do | 12:43 |
| pstolowski | ty | 12:46 |
| zyga | mvo: I'll add my notes in a moment, | 13:16 |
| zyga | I try to add them every day when I EOD | 13:16 |
| mvo | zyga: no worries | 13:23 |
| mvo | zyga: just makes it easier to catchup on details | 13:24 |
| zyga | ack | 13:24 |
| zyga | I can do them before the standup and at EOD next time | 13:24 |
| zyga | my favourite APIs are FooEx | 13:26 |
| zyga | they usually work better than Foo | 13:26 |
| zyga | I wonder if this is how FedEx got the name :P | 13:26 |
| diddledan | is it possible to export the launchpad authentication key for use in CI so I can trigger `snapcraft remote-build` from GitHub Actions? | 13:35 |
| diddledan | my use case is I want more control over which channel I release builds into instead of always going into edge, so I thought to use remote-build in a GitHub Action to get all the architectures built and then in GitHub Actions release them appropriately to alternative channels than edge. | 13:37 |
| diddledan | Github Actions itself is not good for building because it doesn't support ALL the architectures | 13:38 |
| ackk | hi, is there a way for a snap to inspect which slot/plugs it has? | 13:39 |
| ackk | I guess it can read meta/snap.yaml, is there a better way? | 13:39 |
| cjp256 | diddledan: i suppose you could if you constructed the correct credential file for launchpad `$HOME/.local/share/snapcraft/provider/launchpad/credentials` first | 13:42 |
| diddledan | hmm | 13:46 |
| zyga | I opened https://github.com/snapcore/snapd/pull/8389 | 13:47 |
| mup | PR #8389: tests: make session tool way more robust <Test Robustness> <Created by zyga> <https://github.com/snapcore/snapd/pull/8389> | 13:47 |
| zyga | with some luck tests will not fail on anything random in other tests | 13:47 |
| diddledan | don't trust tests that don't fail! | 13:48 |
| diddledan | mind you, you shouldn't trust failing tests, either ;-p | 13:48 |
| diddledan | I _always_ end up having to fix my tests rather than fixing my code because I wrote the test wrong | 13:48 |
| ijohnson | ackk: do you mean connected slots / plugs ? | 13:56 |
| ijohnson | ackk: if so then yes use `snapctl is-connected <name>` | 13:56 |
| === Raboo_ is now known as Raboo | ||
| ackk | ijohnson, no, I mean get the list of names | 13:56 |
| ackk | ijohnson, it seems like looking in snap.yaml would do it, but I was wondering if there was a snapctl way | 13:57 |
| ijohnson | yes, parsing snap.yaml is the best way to just get all the plugs and slots you have | 13:57 |
| ackk | ijohnson, thanks | 13:57 |
| ijohnson | you would have to be slightly careful because you can have for example global plugs in the snap.yaml that apply to every app, and you can have global plugs that are only used in certain places | 13:58 |
| pedronis | mborzecki: did you see this comment https://github.com/snapcore/snapd/pull/8386#discussion_r400830653 ? | 13:58 |
| mup | PR #8386: many: support immediate reboot <UC20> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/8386> | 13:58 |
| mup | PR snapd#8297 closed: tests: session-tool improvements <Created by zyga> <Closed by zyga> <https://github.com/snapcore/snapd/pull/8297> | 13:58 |
| ackk | ijohnson, ah yeah, I was talking specifically about the global ones (like content interfaces) | 13:59 |
| ackk | ijohnson, thanks | 13:59 |
| ijohnson | yes global plugs are bit tricky to work with in figuring out which app/hook they end up applying to | 13:59 |
| ackk | ijohnson, does "is-connected" work for app-specific plugs as well? | 13:59 |
| mborzecki | pedronis: heh, missed it when fixing code | 14:00 |
| ijohnson | ackk: yes `snapctl is-connected` does the right thing depending on which app/snap it was called from and which plugs that app/snap has | 14:00 |
| mborzecki | pedronis: thanks for spotting this | 14:00 |
| ijohnson | ackk: (unless there are bugs of course) | 14:00 |
| ackk | ijohnson, cool, thanks | 14:00 |
| pedronis | mborzecki: apart from that +1 with a suggestion for one of the local vars | 14:01 |
| cjwatson | diddledan: snapcraft remote-build tokens currently expire after a week | 14:18 |
| cjwatson | We might make this more flexible in future, but that's the way it is for now. It was just enough to get going, without having to implement proper token persistence in order that users can revoke them | 14:18 |
| * zyga takes a break to eat lunch | 14:26 | |
| mborzeck1 | omg #8371 has 2 +1, is finally green but the status wasn't reported back to github | 14:27 |
| mup | PR #8371: overlord/devicestate, daemon: record the seed current system was installed from <Needs Samuele review> <UC20> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/8371> | 14:27 |
| === mborzeck1 is now known as mborzecki | ||
| mvo | mborzecki: I can take care of it | 14:30 |
| mborzecki | mvo: i've restarted the cla check, maybe this will update the status | 14:30 |
| mborzecki | mvo: ah you merged it, thanks! | 14:30 |
| mup | PR snapd#8371 closed: overlord/devicestate, daemon: record the seed current system was installed from <Needs Samuele review> <UC20> <Created by bboozzoo> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/8371> | 14:31 |
| mvo | mborzecki: yeah, no need to run spread again and wait forever for silly issues like this, I can just override the failure | 14:32 |
| mborzecki | mvo: travis getting out of sync with gh is so frustrating lately | 14:32 |
| mborzecki | (hope it's not foul play on gh side to promote gh actions though) | 14:33 |
| pedronis | mborzecki: it #8369 ready for re-review by me? | 14:36 |
| mup | PR #8369: boot, overlord/devicestate, daemon: implement requesting boot into a given recovery system <Needs Samuele review> <UC20> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/8369> | 14:36 |
| mborzecki | pedronis: yes it should be | 14:37 |
| mvo | mborzecki: yeah, we need to switch to gh actions all the way | 14:42 |
| zyga | some store woes | 14:42 |
| ijohnson | mborzecki: so I got serial console logs from cachio on a failed uc20 reboot from #8373 | 14:42 |
| mup | PR #8373: tests/lib/prepare.sh: use only initrd from the kernel snap <UC20> <Created by anonymouse64> <https://github.com/snapcore/snapd/pull/8373> | 14:42 |
| ijohnson | see | 14:43 |
| zyga | and I didn't go to eat anything | 14:43 |
| * zyga is gone | 14:43 | |
| ijohnson | https://www.irccloud.com/pastebin/a8UFlkQO/ | 14:43 |
| ijohnson | mborzecki: does this sound familiar? | 14:43 |
| mborzecki | ijohnson: heh, yeah | 14:43 |
| mborzecki | ijohnson: that's what i was talking about ;) | 14:43 |
| ijohnson | so weird | 14:44 |
| mborzecki | ijohnson: i had dumpes of find between the new skeleton and the extracted one, but was not able to conclude anything meaningful out of that | 14:45 |
| ijohnson | so it looks like /main/init in the unpacked initrd skeleton is a symlink to usr/lib/systemd/systemd | 14:45 |
| mborzecki | ijohnson: afaik it shuld be | 14:47 |
| ijohnson | and that's the same in both the distro skeleton and the unpacked skeleton :-( | 14:47 |
| mborzecki | ijohnson: notice that when extrating the initrd tar (?) complains about some symlinks | 14:48 |
| mborzecki | perhaps they should all contain / | 14:48 |
| ijohnson | mborzecki: hmm I don't see that complaint, when do you see that ? | 14:50 |
| ackk | does snapcraft 3.11 support core20-based snaps? | 14:50 |
| mborzecki | ijohnson: i saw it when extracting the initrd, tar was stripping / from symlinks | 14:50 |
| ijohnson | mborzecki: you mean like from inside unmkinitramfs ? | 14:51 |
| ijohnson | I don't see where we are using tar directly anywhere | 14:51 |
| mborzecki | ijohnson: heh no longer there :/ | 15:02 |
| mup | PR snapd#8390 opened: state: add state.CopyState() helper <Created by mvo5> <https://github.com/snapcore/snapd/pull/8390> | 15:04 |
| ijohnson | mborzecki: ah interesting there's: | 15:17 |
| ijohnson | https://www.irccloud.com/pastebin/cmFL9aRT/ | 15:17 |
| ijohnson | but I don't see that when I run the same code on my focal desktop :-/ | 15:17 |
| mborzecki | ijohnson: yup, that's what i saw | 15:17 |
| ijohnson | hmm | 15:17 |
| mborzecki | and yeah, it's cpio not tar | 15:18 |
| ijohnson | ah so there's a new version of initramfs-tools-core available that's not installed in the google images | 15:20 |
| mborzecki | time to wrap it up for today | 15:22 |
| mup | PR snapd#8391 opened: tests: fix cross buidl tests when installing dependencies <Simple 😃> <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/8391> | 15:26 |
| xnox | help snap command fails for me | 15:34 |
| xnox | $ snap info --verbose --arch arm64 pi | 15:34 |
| xnox | error: unknown flag `arch' | 15:34 |
| xnox | how do i get the id of pi snap? | 15:34 |
| pedronis | YbGa9O3dAXl88YLI6Y1bGG74pwBxZyKg | 15:44 |
| ijohnson | ha! https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=946267 is the source of this confusion I bet | 15:45 |
| * ijohnson lunches | 15:45 | |
| xnox | pedronis: right, i have now used obsolete snap-tool that was removed from livecd-rootfs which talks to the store to find that out. | 15:50 |
| xnox | pedronis: i wish either snapcraft or snap commands would support --arch --channel when calling like info or status, to get these things from the store. | 15:50 |
| xnox | setting UBUNTU_STORE_ARCH=arm64 did not help, like it does with `download` command. | 15:50 |
| pedronis | that's been request and is in the backlog but we haven't got to it yet | 15:51 |
| xnox | ok! | 16:00 |
| pstolowski | mvo: hey, what's the target date for 2.44.1? | 16:06 |
| pedronis | pstolowski: you mean .2 ? | 16:09 |
| pstolowski | pedronis: right, yes | 16:10 |
| pedronis | because .1 is going stable now | 16:10 |
| pstolowski | pedronis: yes i mean the next one, with search v2 | 16:12 |
| mvo | pstolowski: about a week from now, why? | 16:13 |
| pstolowski | mvo: Tony was asking | 16:13 |
| pstolowski | thanks | 16:13 |
| mup | PR snapd#8392 opened: tests: remove google-tpm backend from spread.yaml <Simple 😃> <Skip spread> <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/8392> | 16:13 |
| mvo | pstolowski: aha, for searchv2? | 16:15 |
| pstolowski | yes | 16:16 |
| jdstrand | roadmr: hi! can you pull 20200330-2212UTC? note, this has changes that use statefulness for base snaps | 16:37 |
| jdstrand | kenvandine: fyi, that is part of the work needed for the freedesktop base snap ^ | 16:37 |
| zyga | re | 16:41 |
| kenvandine | jdstrand: great | 16:49 |
| roadmr | jdstrand: interesting - yes, I'll queue it up :) | 16:50 |
| jdstrand | roadmr: thanks! | 17:02 |
| ijohnson | cachio: how can I get a debian package updated in the ubuntu-2004-64-virt-uefi-enabled image ? | 17:05 |
| ijohnson | cachio: I need the cpio and initramfs-tools-core debian packages updated in the images | 17:05 |
| cachio | ijohnson, Iwell ubuntu-2004-64-virt-enabled is updated | 17:11 |
| ijohnson | cachio: but the cpio package on that image is out of date | 17:11 |
| cachio | I am cleaning the uefi ones because those are by deault now | 17:11 |
| ijohnson | I have just confirmed this now, it has 2.13+dfsg-1 but it needs 2.13+dfsg-2 | 17:12 |
| cachio | just update the spread.yaml with ubuntu-2004-64-virt-enabled | 17:12 |
| cachio | I think this should work | 17:12 |
| cachio | in that case give me 10 minuttes, I'll update that image | 17:13 |
| ijohnson | cachio: see https://github.com/snapcore/snapd/pull/8373#issuecomment-606758867 | 17:14 |
| mup | PR #8373: tests/lib/prepare.sh: use only initrd from the kernel snap <UC20> <Created by anonymouse64> <https://github.com/snapcore/snapd/pull/8373> | 17:14 |
| ijohnson | cachio: so should I file a PR to change ubuntu-2004-64-virt-uefi-enabled to ubuntu-2004-64-virt-enabled ? | 17:14 |
| cachio | just do it in the one which is failing | 17:15 |
| cachio | to make it quickly | 17:15 |
| ijohnson | cachio: ack will push that up to 8373 | 17:15 |
| cachio | ijohnson, so | 17:17 |
| cachio | google:ubuntu-20.04-64-virt-enabled .../tasks/google/start-instance# apt-cache policy cpio | 17:17 |
| cachio | cpio: | 17:17 |
| cachio | Installed: 2.13+dfsg-2 | 17:17 |
| cachio | Candidate: 2.13+dfsg-2 | 17:17 |
| cachio | so ubuntu-20.04-64-virt-enabled is ok | 17:17 |
| ijohnson | cachio: great | 17:17 |
| ijohnson | cachio: I pushed up that change to the PR, let's see how it does in travis/spread | 17:20 |
| cachio | ijohnson, nice | 17:23 |
| * cachio lunch | 17:38 | |
| mup | PR snapd#8385 closed: tests/main/user-session-env: stop the user session before deleting the test-zsh user <Created by jhenstridge> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/8385> | 17:58 |
| zyga | snapd fails to restore https://www.irccloud.com/pastebin/dPL09dsG/ | 18:46 |
| zyga | leftover LXD mount points https://www.irccloud.com/pastebin/s2YhSMhf/ | 18:47 |
| zyga | cachio: ^ FYI | 18:47 |
| zyga | we should unmount those when we remove lxd | 18:47 |
| zyga | as otherwise we crash in postrm | 18:47 |
| zyga | mvo: ^ LXD mount points are breaking snapd dpkg postrm | 18:48 |
| cachio | zyga, ok, is this caused because of the last change in the lxd snap? | 18:48 |
| zyga | I suspect it's a race | 18:48 |
| zyga | unless they just pushed something | 18:48 |
| zyga | I just saw this for the first time | 18:48 |
| zyga | and I've been running tests all day | 18:49 |
| zyga | but our cleanup step needs to special case lxd | 18:49 |
| zyga | and unmount anything it mounted | 18:49 |
| zyga | which can change at any time | 18:49 |
| cachio | mmmm | 18:50 |
| cachio | ok, in case they change something we are gonna need to make another change | 18:50 |
| cachio | I'll leave some checks to validate we cleaned all the moounts | 18:50 |
| zyga | I'll grab a beer | 18:52 |
| zyga | cachio: we need to fix our postrm sadly | 18:52 |
| zyga | cachio: not only our test stack | 18:53 |
| * zyga is a bit depressed after fighting bugs and compatibility across several releases of various distributions | 18:53 | |
| mvo | zyga: in the tests? | 18:56 |
| zyga | yes | 18:56 |
| zyga | I'm running tests with -repeat | 18:56 |
| zyga | stressing the fixes I made | 18:56 |
| zyga | and I find lots of funky bugs | 18:57 |
| mvo | zyga: is it a real bug that happens when someone removes the lxd snap too? | 18:57 |
| mvo | zyga: or just test internal? | 18:57 |
| zyga | I suspect it depends on how you remove it | 18:57 |
| cachio | zyga, mvo I never saw that in the travis logs | 18:57 |
| zyga | but I suspect apt --purge snapd at the right time can cause that | 18:57 |
| zyga | since we rely on rm -rf stuff | 18:57 |
| zyga | and lxd does create mount points beyond our control | 18:58 |
| pedronis | afaict it doesn't have a remove hook | 19:06 |
| zyga | perhaps it should | 19:07 |
| zyga | to clean up | 19:07 |
| pedronis | anyway in theory we should run remove hooks when we are trying to remove snaps | 19:08 |
| pedronis | wouldn't help here because there isn't one | 19:08 |
| zyga | I agree | 19:35 |
| * zyga EODs depressingly | 19:36 | |
| jdstrand | zyga: sorry you EOD'd depressingly. fyi, I will be doing PR reviews tomorrow, including yours | 19:43 |
| zyga | jdstrand: no worries, it's just ... tough sometimes | 19:48 |
| zyga | jdstrand: my thoughts on this are https://twitter.com/zygoon/status/1245062988358389767 | 19:49 |
| mup | PR snapd#8388 closed: cmd/snap,seed: validate full seeds (UC 16/18) <Squash-merge> <Created by pedronis> <Merged by pedronis> <https://github.com/snapcore/snapd/pull/8388> | 20:19 |
| mup | PR core20#31 closed: hooks/motd: disable dynamic motd, cleanup dangling symlink <Created by bboozzoo> <Merged by xnox> <https://github.com/snapcore/core20/pull/31> | 20:44 |
| jdstrand | zyga: :\ | 20:50 |
| ijohnson | jdstrand: hey did you see my ping from a few months ago on https://bugs.launchpad.net/ubuntu/+source/snapd/+bug/1656340 ? | 21:50 |
| mup | Bug #1656340: XDG_RUNTIME_DIR is not created on app startup <bionic> <cosmic> <eco-team> <xenial> <Snappy:Triaged by zyga> <snapd (Ubuntu):Confirmed> <https://launchpad.net/bugs/1656340> | 21:50 |
| mup | PR snapcraft#3005 closed: Change dotnet runtime version from snapcraft.yaml <Created by amka> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3005> | 22:04 |
| jdstrand | ijohnson: I did. opinions abound and there is no resolution atm. there was a discussion which I captured in trello where I am to sketch out my thought on how to implement things and why. it isn't roadmapped so it is under other things | 22:05 |
| ijohnson | jdstrand: ack that's fair thanks for the heads-up | 22:18 |
| zyga | https://github.com/snapcore/snapd/pull/8389 should go green | 23:08 |
| zyga | hopefully | 23:08 |
| mup | PR #8389: tests: make session tool way more robust <Test Robustness> <Created by zyga> <https://github.com/snapcore/snapd/pull/8389> | 23:08 |
| zyga | it's such a terrible thing, it's still imperfect | 23:08 |
| zyga | but the error rate is lower | 23:08 |
| zyga | and I don't know why it fails (not yet) perhaps just bugs in old systemd/logind | 23:09 |
| zyga | I only saw failures on amazon linux 2 and debian 9 | 23:09 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!