| opendevreview | Maxim Sava proposed openstack/tempest master: Add delete image from store API https://review.opendev.org/c/openstack/tempest/+/898544 | 07:13 |
|---|---|---|
| opendevreview | Maxim Sava proposed openstack/tempest master: Add delete image from store API https://review.opendev.org/c/openstack/tempest/+/898544 | 07:27 |
| ykarel | Hi is there a known issue with test_resize_volume_backed_server_confirm which fails randomly with Kernel panic in guest vm? | 08:13 |
| ykarel | /sbin/init: can't load library 'libtirpc.so.3' Kernel panic - not syncing: Attempted to kill init! exitcode=0x00001000 | 08:13 |
| ykarel | ex:- | 08:13 |
| ykarel | http://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_ff8/898137/1/check/nova-ceph-multistore/ff880e4/testr_results.html | 08:13 |
| ykarel | https://d941f4f9ec28b33784a2-4d240979ebaa10fc274be1cd05b244a9.ssl.cf5.rackcdn.com/893025/3/check/tempest-integrated-networking/7ed5444/testr_results.html | 08:14 |
| ykarel | as per opensearch there are 16 hits across multiple jobs | 08:14 |
| ykarel | gibi, sean-k-mooney dansmith in case you know ^? | 08:15 |
| ykarel | Reported https://bugs.launchpad.net/tempest/+bug/2039940 to track | 08:30 |
| sean-k-mooney | ykarel: not that i know of specificaly but there are some intermint kernel crashes that we see. just not specificaly corralated to any one test | 10:20 |
| ykarel | sean-k-mooney, ack one other kernel panic that i am aware off is related to IO apic and cirros qcow images, not sure if you referring those | 10:28 |
| sean-k-mooney | yes the io apic issue was related to the ubuntu 18.04/20.04 kernel in cirros 5 | 10:38 |
| sean-k-mooney | it was fixed in a later lts kernel but cirros never rebuilt with it | 10:38 |
| sean-k-mooney | cirros 6 shoudl not hit that io apic issue | 10:38 |
| sean-k-mooney | well unless there is a second kernel bug in the newer kernel | 10:38 |
| ykarel | ack in neutron jobs we were using cirros partitioned images to workaround that IO apic issue | 10:40 |
| sean-k-mooney | huh i was not aware of that workaround | 11:20 |
| sean-k-mooney | odd that partioned vs not would have any delta in behavior | 11:20 |
| ykarel | sean-k-mooney, by workaround for those partitioned images i meant https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L6129-L6132 | 12:00 |
| sean-k-mooney | ah you use the split out kernel image and aset the commandline arg | 12:02 |
| sean-k-mooney | wehn we first hti this we agreed that we shoudl not convert all jobs to that approch because it not the way vms are normally booted and we didnt want to lose coverge for the normal codepath | 12:03 |
| sean-k-mooney | so i hope you dont have that set in all jobs | 12:03 |
| ykarel | yes not using in all the jobs | 12:04 |
| sean-k-mooney | i need to revive my alpine seise at some point | 12:04 |
| sean-k-mooney | https://review.opendev.org/q/topic:alpine that was a way we hoped to mitigate this, use alpine instead of cirros | 12:05 |
| frickler | if the issue is really due to the way cirros images work starting out with essentially just an initramfs, pretty sure smoser and me can be convinced to also provide another image type. the current setup was what was invented to avoid a lot of duplication in the image, but if it turns out to cause issues, we sure want to deal with them | 12:25 |
| frickler | and nowadays 50% of the image size is just the kernel anyway | 12:25 |
| dansmith | frickler: kopecmartin: can either of you +W this? https://review.opendev.org/c/openstack/grenade/+/898745 | 13:45 |
| dansmith | still hitting that problem in our queues and I'm hoping that will alleviate it | 13:45 |
| dansmith | tl;dr: I've seen some instantaneous, some 30<t<60 and some slightly over 60s syncs, but no total lockups, which means to me that waiting longer makes sense | 13:46 |
| opendevreview | Luigi Toscano proposed openstack/devstack-plugin-nfs master: zuul: run also tests from cinder-tempest-plugin https://review.opendev.org/c/openstack/devstack-plugin-nfs/+/898965 | 14:11 |
| opendevreview | Luigi Toscano proposed openstack/devstack-plugin-nfs master: zuul: run also tests from cinder-tempest-plugin https://review.opendev.org/c/openstack/devstack-plugin-nfs/+/898965 | 14:51 |
| opendevreview | Luigi Toscano proposed openstack/devstack-plugin-nfs master: zuul: run also tests from cinder-tempest-plugin https://review.opendev.org/c/openstack/devstack-plugin-nfs/+/898965 | 15:00 |
| sean-k-mooney | frickler: the io-apic thing was a kernel bug unrelated to how the image was pacakged but you could pass kernel args to work aroudn it if you used a seperate kernel image | 15:04 |
| sean-k-mooney | frickler: os in principal just doing an package update in 5.x woudl have been suffinct but when i ask about htat a few years ago it never happend. and since then we have had 6.x which shoudl be patch so its not as much of a priority. | 15:05 |
| opendevreview | Merged openstack/grenade master: Increase cinder verify sync timeout https://review.opendev.org/c/openstack/grenade/+/898745 | 15:57 |
| opendevreview | James Parker proposed openstack/whitebox-tempest-plugin master: Update requirements for paramiko for Downstream Train https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/898747 | 16:39 |
| opendevreview | James Parker proposed openstack/whitebox-tempest-plugin master: Update requirements for paramiko for Downstream Train https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/898747 | 16:42 |
| opendevreview | James Parker proposed openstack/whitebox-tempest-plugin master: Update requirements for paramiko for Downstream Train https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/898747 | 16:46 |
| opendevreview | Merged openstack/whitebox-tempest-plugin master: Update requirements for paramiko for Downstream Train https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/898747 | 22:27 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!