| === wxl is now known as NotAGuest | ||
| === NotAGuest is now known as wxl | ||
| mwhudson | vorlon: thanks for looking at the ubuntu-cdimage stuff | 00:09 |
|---|---|---|
| mwhudson | vorlon: when you say "We'll have to subject this to testing to figure out in the end if it's correct." do you mean before landing? | 00:10 |
| vorlon | mwhudson: I mean that landing it and making images that get community testing is the only way to figure out if e.g. we've fixed LP: #1886148 | 00:53 |
| ubottu | Launchpad bug 1886148 in Ubuntu CD Images "failure to boot groovy daily" [Undecided,In progress] https://launchpad.net/bugs/1886148 | 00:53 |
| mwhudson | vorlon: yeah ok | 00:53 |
| mwhudson | makes sense to me | 00:53 |
| vorlon | lol I guess I should update my debian-cd bzr checkout to stop pushing to nusakan | 00:59 |
| mwhudson | vorlon: bonus points for merging on my friday | 01:10 |
| vorlon | hah | 01:20 |
| slyon | I have some problems with autopkgtests, using the "needs-reboot" restriction and triggering a reboot during a (new) test via "/tmp/autopkgtest-reboot", introduced in "netplan.io 0.100-0ubuntu2". | 07:42 |
| slyon | On ppc64el the test seems to be spinning in a loop (started over and over again) since 2 days. While for amd64 and arm64 the test fails with a broken SSH connection to nova "<VirtSubproc>: failure: Timed out on waiting for ssh connection". | 07:42 |
| slyon | I think somebody must manually cancel the ppc64el infinite loop... Also I'm wondering if there are other tests using the "needs-reboot" restriction, which I could compare my new test to (which is working just fine in my local qemu setup)? | 07:43 |
| cpaelzer | slyon: https://autopkgtest.ubuntu.com/packages/d/dbus/groovy/amd64 is an example | 08:17 |
| cpaelzer | or any of https://codesearch.debian.net/search?q=Restrictions.*needs-reboot&literal=0 | 08:18 |
| cpaelzer | the ppc test is only running since 2h 27m 49s | 08:19 |
| cpaelzer | only submit date is from 2020-09-09 15:12:50 | 08:19 |
| slyon | cpaelzer: thanks I'll look at those and compare what they're doing differently. | 08:24 |
| slyon | The ppc test was uploaded 2020-09-09 and I'm observing it since then. I think it always runs for a couple of ours, until the last (reboot) test appears. Then it (probably) reboots, resets the time and starts over. I've seen it executing the same tests over and over again since the last 2 days | 08:25 |
| slyon | *couple of hours | 08:25 |
| aleb | Until when can we make a new Pitivi release https://packages.ubuntu.com/groovy/video/pitivi so it's included in Groovy? | 08:30 |
| cpaelzer | aleb: as-is until some weeks ago :-/ | 08:33 |
| cpaelzer | aleb: you'll need https://wiki.ubuntu.com/FreezeExceptionProcess | 08:33 |
| cpaelzer | aleb: release schedule is here https://discourse.ubuntu.com/t/groovy-gorilla-release-schedule/15531 | 08:33 |
| aleb | I saw the release schedule. Ok, I'll file a bug in launchpad after we release. Thanks! | 08:34 |
| === guiverc2 is now known as guiverc | ||
| cpaelzer | while that is the right way - to be clear the remaining schedule is tight, no promises that everything is ready and accepted | 08:35 |
| cpaelzer | aleb: if I might ask, will that be a 1.0 then? | 08:36 |
| cpaelzer | as it is 0.99 into 0.999 so far? | 08:36 |
| aleb | The "1.0" turned into focusing that the GES backend is stable, and the app is as stable as possible. We do have lots of tests also for the app (UI). The decision of not including new features just to keep the app stable was not good for the health of the project though. We were staying with the ass in two boats, as the saying goes. We'll include lots of features introduced in GSoC projects the past years. The next releas | 08:45 |
| aleb | will be "20.09" etc, we'll never release "1.0". | 08:45 |
| === bashfulrobot is now known as Guest25928 | ||
| cpaelzer | doko: I was seeing a few more armhf gcc-10 segfaults like https://launchpadlibrarian.net/497142982/buildlog_ubuntu-groovy-armhf.dpdk_20.08-1ubuntu1~ppa1_BUILDING.txt.gz | 10:27 |
| cpaelzer | doko: is there anything already in the queue that could adress this? | 10:27 |
| cpaelzer | I added it to 1890435 but I can't reliably recreate it yet | 10:30 |
| cpaelzer | so not much new isight | 10:30 |
| cpaelzer | insight | 10:30 |
| jibel | bdmurray, hi, there is a string fix needed for https://code.launchpad.net/~brian-murray/ubiquity/+git/ubiquity/+merge/390185 then I'll merge it as soon as you pushed the change. | 12:44 |
| coreycb | icey: about python3-dnspython 2.0.0 breaking python3-eventlet in groovy | 13:31 |
| coreycb | I think this summarizes most of the known issues? https://github.com/eventlet/eventlet/issues/619 | 13:32 |
| icey | coreycb: https://bugs.launchpad.net/designate/+bug/1895285 is a different dnspython2 issue | 13:33 |
| ubottu | Launchpad bug 1895285 in Designate "Designate doesn't work with dnspython 2.0.0" [Undecided,New] | 13:33 |
| icey | but yes, with regards to eventlet, that's a good bug | 13:33 |
| coreycb | upstream openstack tried to move to 2.0.0 but had to move back to 1.16.0 due to all the issues with eventlet + dnspython 2.0.0 - https://review.opendev.org/#/c/741986/ | 13:36 |
| coreycb | icey: ^ | 13:36 |
| coreycb | icey: upstream is actively working on a fix but at this point in groovy I don't see this getting resolved, and openstack heavily uses eventlet unfortunately | 13:40 |
| icey | indeed | 13:41 |
| slyon | cpaelzer: I think I found the root cause for the SSH timeout issue in my autopkgtest: The management network interface seems to be down after playing with netplan during the test. I set up a local qemu autopkgtest VM, which is accessed via the autopkgtest-virt-ssh runner and was able to "reporduce" the problem. (Well... its not actually on nova/openstack). | 14:07 |
| slyon | cpaelzer: I came up with this debdiff, which fixes the problem in my local "fake ssh runner": https://paste.ubuntu.com/p/46VrKmfH6Z/ | 14:08 |
| slyon | is there any way how to test that in the real nova/openstack environment, without uploading it to the archive? | 14:08 |
| bdmurray | jibel: updated, thanks | 14:11 |
| ahasenack | hi, is there an easy way to inject a ppa into debootstrap, so it considers the packages from the ppa in addition to the ones from the archive? | 14:21 |
| ahasenack | I guess via the script argument? | 14:21 |
| ahasenack | or hacking /usr/share/debootstrap/scripts/groovy (in the groovy case) directly? | 14:23 |
| ahasenack | I need to test the first-time installation of base-files | 14:35 |
| ahasenack | without uploading it to groovy first | 14:36 |
| sem2peie | good evening, is a developer present who deals with linux-image-generic-hwe-20.04 / ath10k and WLAN cards? | 17:57 |
| sem2peie | I have created a ticket, maybe someone can take a look https://bugs.launchpad.net/ubuntu/+bug/1895333 | 18:07 |
| ubottu | Launchpad bug 1895333 in Ubuntu " cannot change the regulatory domain ath10k, QCA9984 (QNAP QWA-AC2600)" [Undecided,New] | 18:07 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!