[07:07] morning [08:02] good morning [08:07] morning [08:09] hey pstolowski [08:18] zyga: pstolowski: hey guys [09:04] heh, so another bunch of rebuilds and had to go retrigger the builds on obs [09:04] i guess it's time i tweaked those client timeouts [09:09] PR snapd#10059 opened: cmd/snap: use less aggressive client timeouts in unit tests [10:04] PR snapd#10060 opened: boot: extend set try recovery system unit tests [10:12] mborzecki: hi, should we close 100055, as it's replaced by 100056 ? [10:16] pedronis: yeah, closed 10055 just now [10:16] thx [10:16] pedronis: shall we sync with mvo/ijohnson about the issue today? [10:16] mvo is still off [10:16] but yes we should sync [10:16] ah [10:19] PR snapd#10055 closed: wrappers: launch apps through /usr/lib/snapd/snap on core with snapd snap <⛔ Blocked> [10:33] mborzecki: we can chat before or after the standup [10:33] pedronis: let's try before [10:34] pstolowski: hi, maybe you could review https://github.com/snapcore/snapd/pull/10040 ? [10:34] PR #10040: daemon: switch api_test.go to daemon_test and various other cleanups [10:34] pedronis: sure [10:44] pedronis: i've updated https://github.com/snapcore/snapd/pull/10037 [10:44] PR #10037: boot: export helper for clearing tried system state, add tests [10:53] mborzecki: thanks, +1 [10:53] thanks for the review [11:11] alan_g: I proposed some more fixes to help with #8699 here: https://github.com/MirServer/snapd/pull/5 -- there is still a bit more to address, but this should resolve most of the open threads [11:11] PR #8699: interfaces/desktop-launch: support confined snaps launching other snaps [11:11] PR MirServer/snapd#5: usersession/userd: fix some of the review comments for #8699 [11:12] jamesh, thanks. Will take a look [11:14] alan_g: I did a merge from master though, which complicates the diff. The changes in usersession/userd should be relatively clean though [11:14] jamesh, Ack. I spotted that already [11:50] PR snapcraft#3480 opened: ci: spread testing [12:00] * cachio afk [12:00] PR snapcraft#3480 closed: ci: spread testing [12:02] mborzecki: I scheduled something [12:02] thanks! [12:15] PR snapcraft#3481 opened: tests: stub job to get autokpgtest for edge === jamesh_ is now known as jamesh [13:05] cachio: hey it seems we may have hit our image limit again, I see `2021-03-22 04:12:51 Cannot allocate google-nested:ubuntu-18.04-64: cannot find any Google image matching "ubuntu-1804-64-virt-enabled" on project "computeengine" or "ubuntu-os-cloud"` [13:05] can you clean up some images on gCE? [13:17] ijohnson, hi [13:17] let me check [13:27] thanks [13:28] ijohnson, it is done [13:28] try again please [13:28] great, thanks I will run it again [14:20] PR snapd#10061 opened: interfaces/pwm: add PWM interface [15:10] PR snapcraft#3482 opened: Support install-mode option for apps [15:29] pedronis , ijohnson : i've pushed the version using findmnt to https://github.com/snapcore/core20/pull/100 [15:29] PR core20#100: Add a wrapper for invoking the snap binary [15:31] mborzecki: great I'll have a look in a bit [15:37] mborzecki: it's better :) [15:40] pedronis: i'll force push that || -> && [16:17] pstolowski: I did a first on affectedByRefresh PR [16:17] *pass [16:20] pedronis: thank you [16:21] PR snapd#10062 opened: tests: validation sets spread test [16:36] PR snapd#10063 opened: o/{device,hook}state: encode fde-setup-request key as base64 string <⚠ Critical> [16:42] ijohnson: hey, do you have a moment to do 2nd review of https://github.com/snapcore/snapd/pull/10048 ? [16:42] PR #10048: o/configstate: deal with no longer valid refresh.timer=managed [16:44] pstolowski: sure I will add it to my queue [16:44] thanks === ijohnson is now known as ijohnson|lunch [18:06] ijohnson|lunch, hi, did you start taking a look to the auto-refresh test? === ijohnson|lunch is now known as ijohnson [18:09] cachio: sorry can you remind me which test? [18:09] ijohnson, auto-refresh [18:10] cachio: you mean tests/main/auto-refresh ? [18:10] tes [18:10] yes [18:17] ijohnson, the snap refresh --time is showind a different output [18:17] https://paste.ubuntu.com/p/ykkrp3Xcqc/ [18:18] dependening if we run in the first test suite or in the second one [18:18] the test suite clean up it messing up something I guess [18:19] cachio: the test runs ` snap set core refresh.schedule="0:00-23:59"` [18:20] cachio: the test needs to do in the restore sectoin `snap unset core refresh.schedule` [18:21] the problem is before taht [18:22] first thing the test does is [18:22] echo "Auto refresh information is shown" [18:22] output=$(snap refresh --time) [18:22] well [18:22] cachio: but what you are saying is that the `snap refresh --time` fails in the _second_ instance of running the test right ? [18:22] ijohnson, no [18:22] it fails at the beginning [18:22] I guess I'm confused how the test is failing then [18:23] also confused when the test is failing [18:23] in fact I shell the test and I see the diferent output [18:23] if I run [18:23] spread2 -debug -order -single-worker google:arch-linux-64:tests/completion/indirect:funkyfunc google:arch-linux-64:tests/main/auto-refresh:parallel [18:23] so then the output for nap refresh --time [18:24] contains timer: 00:00~24:00/4 [18:24] if I run [18:24] spread2 -debug -order -single-worker google:arch-linux-64:tests/completion/indirect:funkyfunc google:arch-linux-64:tests/main/auto-refresh:parallel [18:24] spread2 -debug -order -single-worker google:arch-linux-64:tests/main/auto-refresh:parallel [18:24] last one [18:24] so then the output for nap refresh --time [18:24] contains schedule: 00:00-23:59 [18:25] hmm [18:25] so the test fails not when the auto-refresh test is run twice, but when any other test is run before it ? [18:25] another test in another suite [18:25] yes [18:26] I see [18:26] I ran in a loop and the test works well [18:26] I wonder what the restore logic is doing [18:26] any idea what could be affecting this? [18:26] cachio: where do we have the magical code that resets snapd state that is run after every spread test? [18:27] tests/lib/reset.sh [18:31] ijohnson, after the last change the reset houldn't be needed [18:31] because we reaet after running the last test [18:33] cachio: when you see the auto-refresh test fail, can you get an output of `snap get system` ? [18:33] ijohnson, https://paste.ubuntu.com/p/s7ccZcPHdY/ [18:35] cachio: sorry `snap get system -d` [18:36] ijohnson, https://paste.ubuntu.com/p/pn2MD2FSSR/ [18:36] cachio: something set `snap set core refresh.hold=...` on that system [18:38] cachio: can you check what cloud-init scripts / datasources are doing for the system [18:38] ? [18:38] maybe someone added to the image something which calls `snap set core refresh.hold=...` [18:38] let me check [18:45] ijohnson, cloud init is not [18:48] cachio: you mean cloud-init is not running `snap set core refresh.hold=` ? [18:49] also cachio can you please review https://github.com/snapcore/snapd/pull/10044 and +1 if it it makes sense to you ? [18:49] PR #10044: packaging: drop dh-systemd from build-depends on ubuntu-16.04+ <⚠ Critical> [18:50] ijohnson, sure [18:52] ijohnson, cloud init is not setting snap set core refresh.hold= [18:52] hmm [18:52] I'll continue working on this [18:52] don't worry [18:56] ok, thanks I don't have much bandwidth for this auto-refresh thing right now, and I think it's more pressing for me to get you a fix for the uevent spread test instead of debugging this [18:56] PR snapd#10064 opened: tests: simplify the reset.sh logic by removing not needed command [18:56] but obviously feel free to reach out to others tomorrow morning about auto-refresh maybe someone else has an idea about what is going on [19:04] ijohnson, sure, thanks a lot [20:51] PR snapcraft#3483 opened: requirements: use PyNaCl 1.3.0 and ensure is compiled on linux [21:47] * cachio afk [21:51] PR snapcraft#3484 opened: tests: crystal 1.0.0 requires shard.lock [21:52] PR snapd#10065 opened: tests: reset the system while preparing the test suite [22:41] PR snapcraft#3484 closed: tests: crystal 1.0.0 requires shard.lock