Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-09-30
13:42:18 lyarwood I'll add a comment
13:43:53 dansmith ah, this guy: https://review.opendev.org/#/c/752557/
13:44:25 lyarwood that's the same change as above
13:44:40 dansmith lol
13:45:25 lyarwood https://review.opendev.org/#/c/742056/ is the grenade change
13:45:26 dansmith I know, I just didn't realize that was yours, thought it was gmann's because he had the -W until a minute ago
13:45:38 lyarwood ah np
13:45:46 dansmith yeah, I got it now
13:46:12 dansmith btw, I'm super happy to see you guys also getting spanked by zuul on syntax stuff
13:46:33 dansmith fixing up glance's jobs, zuul was getting sick of me throwing up patches that didn't even parse :)
13:46:55 lyarwood haha yeah, all the damn time
13:47:05 lyarwood wish there was a way of running some of this stuff locally
13:47:08 dansmith there needs to be a separately-installable zuulint tool or something we can throw into a tox env
13:47:11 dansmith yeah
14:02:09 sean-k-mooney lyarwood: there is a docker compose file for running zull quickly
14:02:18 sean-k-mooney but ya gettign the gate jobs to work is non trivial
14:03:16 sean-k-mooney the devstack/tempest based jobs pull in a lot of repos so you need to donload a lot of repos to get them to work unless you take a few steps that mean your nolonger really mirrioring what happens in the gate
14:04:00 sean-k-mooney dansmith: but ya a zuul-lint tool would be nice
14:47:36 openstackgerrit Lee Yarwood proposed openstack/nova master: WIP zuul: Replace nova-live-migration with zuulv3 jobs https://review.opendev.org/752557
14:47:36 openstackgerrit Lee Yarwood proposed openstack/nova master: WIP zuul: Merge nova-evacuate into nova-multinode-live-migration https://review.opendev.org/755309
17:46:32 openstackgerrit Lee Yarwood proposed openstack/nova stable/victoria: [stable-only] nova-live-migration: Disable libvirtd during negative tests https://review.opendev.org/754688
17:49:07 openstackgerrit melanie witt proposed openstack/nova master: Add note about setting CONF.max_disk_devices_to_attach = 0 https://review.opendev.org/755360
17:56:27 openstackgerrit melanie witt proposed openstack/nova master: Add note about setting CONF.max_disk_devices_to_attach = 0 https://review.opendev.org/755360
20:11:08 ganso Hi Nova folks! I know most people are very busy with bugs for the RC1, but just in case you have 1 minute to spare, here is a backport to stable/ussuri that is waiting for +2s, if you could please take a look at. Thanks in advance! https://review.opendev.org/#/c/752245/
20:31:32 sean-k-mooney lyarwood: elod ^ you might be able to help ganso when ye are around
21:58:14 ganso sean-k-mooney: thanks!
#openstack-nova - 2020-10-01
08:05:21 hemanth_n Hi folks, is this time the unofficial office hours slot , if so want to discuss on the following review https://review.opendev.org/#/c/749175/
08:11:16 lyarwood hemanth_n: I don't think we have any office hours at the moment but looking at the change gibi / sean-k-mooney1 / stephenfin might be able to help once they are online
08:11:42 lyarwood In other news, stable/victoria is blocked until we switch the nova-live-migration job out :/
08:11:45 lyarwood ++ /opt/stack/new/devstack-gate/devstack-vm-gate.sh:setup_localrc:L316: /usr/bin/python3 roles/test-matrix/library/test_matrix.py -n -b stable/victoria -f roles/test-matrix/files/features.yaml -r primary
08:11:47 lyarwood 2020-09-30 21:04:43,670 ERROR: branch not allowed by features matrix: victoria
08:12:32 hemanth_n thanks lyardwood
08:13:02 hemanth_n lyarwood* thank you
08:13:27 lyarwood np
08:13:52 gibi hemanth_n: that patch is on my review list for long
08:14:57 gibi hemanth_n: I see that sean-k-mooney1 is +1 on it already so I will try to read it today
08:15:08 gibi hemanth_n: is there any special aspect I should be aware of?
08:15:53 hemanth_n gibi: cool thanks .. nope except its a specific scenario which is mentioned in the commit message and bug description
08:17:07 gibi hemanth_n: I saw that you managed to add a functional test for it. thanks that helps a lot
08:17:32 openstackgerrit Lee Yarwood proposed openstack/nova stable/victoria: [stable-only] nova-live-migration: Disable libvirtd during negative tests https://review.opendev.org/754688
08:25:09 frickler lyarwood: the right solution would be to move that job away from using ds-gate, I'm not sure if someone has started doing that yet. even if we keep ds-gate working for victoria, that should be considered unstable and very much deprecated
08:25:56 lyarwood frickler: yeah I'm working on that in master, just need sort out a multinode ceph job and we should be good to backport to stable/victoria
08:26:09 lyarwood frickler: and agreed, this is just for a few days hopefully
08:27:26 lyarwood gibi: are you aware of https://zuul.opendev.org/t/openstack/builds?job_name=nova-next - looks like gate/post_test_hook.sh is failing when it calls `openstack port unset --binding-profile allocation port-normal-qos`
08:27:41 lyarwood on master at least
08:28:02 gibi lyarwood: thanks for the ping
08:28:10 gibi I haven't seen it yet
08:29:50 lyarwood gibi: np, can't see anything obvious in nova master that could be causing it, I assume something in Neutron has changed?
08:32:22 gibi lyarwood: I will ping lajoskatona on Neutron side, he did some recent changes around qos
08:33:28 lyarwood awesome thanks
08:35:12 openstackgerrit Lee Yarwood proposed openstack/nova stable/victoria: [stable-only] nova-live-migration: Disable libvirtd during negative tests https://review.opendev.org/754688
09:10:07 openstackgerrit Lee Yarwood proposed openstack/nova master: WIP: nova-next: Start testing the 'q35' machine type https://review.opendev.org/708701
10:17:09 stephenfin gibi: Think you could spin through https://review.opendev.org/#/q/topic:remove-xenapi+status:open before the week is out?
10:24:53 gibi stephenfin: I'm not comfortable making big changes to master during release time, can we take this after the release?
10:25:06 stephenfin oh, of course
10:30:34 openstackgerrit Balazs Gibizer proposed openstack/nova-specs master: Support interface attach with qos ports https://review.opendev.org/755477
11:13:34 openstackgerrit Stephen Finucane proposed openstack/nova-specs master: Add modernize-os-hypervisors-api spec https://review.opendev.org/755109
11:18:10 gibi lyarwood: it seem neutron-lib needs a new release to contain this: https://review.opendev.org/#/c/753230/
11:18:56 lyarwood gibi: ack thanks for looking into that!
11:19:18 lyarwood gibi: ./me wonders if neutron have nova-next in their check/gate?
11:19:28 gibi they don't I guess
11:19:44 lyarwood yeah might be something to suggest given the QoS coverage it has
11:22:02 gibi good point
11:22:35 gibi to be fair the normal qos coverage is in tempest. the nova-next only has the coverage for the qos part of the nova-manage placement heal_allocation CLI
11:23:20 lyarwood ah so we could add something to tempest to also cover this?
11:26:45 gibi I think it was added to the nova-next job as it has to create an invalid situation first then heal it
11:27:02 gibi in a real word missing port allocation only happen during upgrade from older openstack version
11:27:13 gibi we cannot really simulate that
11:27:26 gibi so we instead deleteing the allocation manually
11:27:29 gibi then try to heal it
11:29:20 lyarwood ack
11:30:03 lyarwood still would be nice if a neutron job could execute the test somewhere in their gate
11:30:17 lyarwood if we could break the test out into a seperate script somewhere
11:30:37 gibi yeah, I agree with that
11:37:02 gibi lyarwood: I guess the whole nova gate is broken now due to this
11:43:14 gibi lyarwood: I'll try to disable this part of the nova-next job as it will take time to get a new neutron-lib release
11:55:29 gibi lyarwood: also stable/victoria could have the same issue if https://review.opendev.org/#/c/755180 merges
12:05:46 openstackgerrit Balazs Gibizer proposed openstack/nova master: Temporarily disable parts of heal port allocation test https://review.opendev.org/755498
12:05:52 gibi lyarwood: ^^
12:07:49 openstackgerrit Balazs Gibizer proposed openstack/nova master: Temporarily disable parts of heal port allocation test https://review.opendev.org/755498
12:15:39 openstackgerrit Balazs Gibizer proposed openstack/nova master: Temporarily disable parts of heal port allocation test https://review.opendev.org/755498
12:15:52 lyarwood gibi: ack thanks
12:16:18 gibi slaweq and lajoskatona are aware of the issue and working on a solution
12:17:03 lyarwood awesome thanks
12:17:27 lyarwood stephenfin: ^ the above should unblock the gate on master btw
12:23:02 openstackgerrit Rodrigo Barbieri proposed openstack/nova stable/ussuri: Allow tap interface with multiqueue https://review.opendev.org/752245
12:23:13 openstackgerrit Rodrigo Barbieri proposed openstack/nova stable/ussuri: Allow tap interface with multiqueue https://review.opendev.org/752245
12:31:07 openstackgerrit Elod Illes proposed openstack/nova stable/ussuri: Allow tap interface with multiqueue https://review.opendev.org/752245
13:09:08 openstackgerrit Hemanth Nakkina proposed openstack/nova master: Update pci stat pools based on PCI device changes https://review.opendev.org/749175
13:14:04 openstackgerrit Lee Yarwood proposed openstack/nova master: WIP: nova-next: Start testing the 'q35' machine type https://review.opendev.org/708701
13:14:16 lyarwood kashyap: ^ think I found a n-cpu bug with QEMU v5.0.0, looks like we will need to rewrite our detach logic
13:14:44 lyarwood I'll write up a bug once the vanilla fedora-latest-virt-preview job has also hit it without q35
13:20:04 sean-k-mooney lyarwood: is this related the the qemu issue
13:20:08 sean-k-mooney on ubuntu
13:20:26 sean-k-mooney i.e. would the same rewite help
13:21:02 lyarwood sean-k-mooney: it's unrelated
13:22:50 lyarwood sean-k-mooney: https://github.com/qemu/qemu/blob/cbba3dc6ea3fc9aa66e9f9eb41051536e3ad7cd0/qdev-monitor.c#L886-L898 - n-cpu is tripping up over this check introduced in QEMU v5.0.0 as we fire off detach requests without waiting for libvirt to emit events.
13:23:20 lyarwood hmm unless
13:23:44 sean-k-mooney if (dev->pending_deleted_event) {
13:23:45 lyarwood are we hitting the bug this check was introduced to avoid?

Earlier   Later