Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2021-02-08
14:36:20 gibi if there is no host that is in both agg2 and agg3 or agg1 and agg3 then yes it is a no valid host
14:36:29 gibi but a host can be in multiple aggregated
14:36:32 gibi -d
14:36:44 bauzas ok, b/c my functest only has one host per agg
14:37:00 bauzas but I can try to do something
14:37:20 bauzas like, adding host5 that's in both aggregates
14:37:24 bauzas lemme try about it
14:38:56 gibi ok
14:54:58 sean-k-mooney stephenfin: sorry was on a call that sounds fine by me
14:55:28 sean-k-mooney stephenfin: i was just inspecting some data in the db for one of my changes and found it odd that those tabels were still there
15:00:29 bauzas I'm absolutely puzzled
15:00:42 bauzas looks like our HostNameWeigherFixture doesn't work
15:01:25 gibi bauzas: that is only handles host1,2,3 in the baseline
15:01:41 bauzas gibi: http://paste.openstack.org/show/802441/
15:01:51 bauzas gibi: aaaaaaah
15:02:10 bauzas so we need to fix this :p
15:02:20 gibi you can pass in weights to the fixture
15:02:26 bauzas i was thinking the weigher was smarter :)
15:02:37 bauzas will lookup
15:02:48 bauzas tbh, I haven't looked at the fixture yet
15:02:53 bauzas RTFM
15:03:00 gibi it is a good enough fixture :)
15:03:05 bauzas or rather, RTFM
15:03:12 bauzas (read the f*** module)
15:03:59 bauzas ok, I see the init call
15:04:09 bauzas we can pass the weights
15:04:14 bauzas all good, thanks for the quick resolution
15:04:39 gibi happy to help
15:10:46 gmann sean-k-mooney: openstack-tempest-skiplist is under/from tripleO project. not new i think
15:11:14 gmann stephenfin: yeah, most of functional tests also run with permission to all (""). we might need more work there too
15:11:38 gmann gibi: ah, I missed that. will work on this today
15:11:46 gibi gmann: thanks
15:14:20 sean-k-mooney gmann: yep its not. i was just worreid if this was being used sytemwide for all tempest jobs
15:15:01 sean-k-mooney gmann:that would be bad since its disiableing things based on wether its supported or not by ooo but since its only used by ooo its fine
15:15:18 gmann sean-k-mooney: no, I think that may be used in tripleO jobs only
15:15:20 sean-k-mooney i just would have expected to see ooo/tripleo in the repo name
15:16:00 gmann yeah that is confusing even last month release team also had same confusion and nobody know if it is Tempest repo or something else
15:16:37 gmann and skip list could be maintained in file itself, I do not think separate repo is needed
15:17:37 sean-k-mooney gmann: ya i think they were not thinking baout zuul job inheritance wehn they created it but if it works for them i guess its ok
15:18:02 sean-k-mooney i just wanted to make sure a patch to that would not skip something in a nova job which it wont so all good
15:56:54 openstackgerrit Lee Yarwood proposed openstack/nova master: libvirt: Record the machine_type of instances in system_metadata https://review.opendev.org/c/openstack/nova/+/767533
15:57:00 openstackgerrit Lee Yarwood proposed openstack/nova master: nova-manage: Add commands for managing instance machine type https://review.opendev.org/c/openstack/nova/+/769548
15:57:04 openstackgerrit Lee Yarwood proposed openstack/nova master: WIP nova-status: Add hw_machine_type check for libvirt instances https://review.opendev.org/c/openstack/nova/+/770643
16:33:20 dansmith stephenfin: ack on the networks thing, I've seen that one too
16:38:29 openstackgerrit Sylvain Bauza proposed openstack/nova master: Add net and utils methods for getting routed netks and segments https://review.opendev.org/c/openstack/nova/+/773976
16:38:29 openstackgerrit Sylvain Bauza proposed openstack/nova master: Add a routed networks scheduler pre-filter https://review.opendev.org/c/openstack/nova/+/749068
16:38:48 bauzas gibi: updated, thanks for your comments ^
16:39:00 bauzas added both functest and UT for verifying it
16:41:40 gibi bauzas: on it
16:52:10 lyarwood melwitt: https://review.opendev.org/c/openstack/nova/+/766976 - before I forget, would you mind hitting this today? You reviewed the bugfix on top of it but missed the regression test on Friday.
17:04:36 openstackgerrit Balazs Gibizer proposed openstack/nova master: libvirt: Define and emit DeviceRemovedEvent and DeviceRemovalFailedEvent https://review.opendev.org/c/openstack/nova/+/749929
17:04:50 gibi stephenfin: fixed your comment in ^^
17:06:14 openstackgerrit Balazs Gibizer proposed openstack/nova master: libvirt: add AsyncDeviceDetachEventsHandler https://review.opendev.org/c/openstack/nova/+/772381
17:06:15 openstackgerrit Balazs Gibizer proposed openstack/nova master: libvirt: allow querying devices from the persistent domain https://review.opendev.org/c/openstack/nova/+/772383
17:09:14 openstackgerrit Balazs Gibizer proposed openstack/nova master: libvirt: parse alias out from device config https://review.opendev.org/c/openstack/nova/+/772384
17:10:38 openstackgerrit Balazs Gibizer proposed openstack/nova master: Replace blind retry with libvirt event waiting in detach https://review.opendev.org/c/openstack/nova/+/770246
17:16:38 gibi bauzas: I'm +2 on the routed net series, good job!
17:16:47 bauzas thanks
17:17:01 gibi sean-k-mooney: I would appreciate your view on the question in https://review.opendev.org/c/openstack/nova/+/749068/12/nova/scheduler/request_filter.py#312
17:17:25 bauzas gibi: then hopefully I'll switch soon on the compute RPC API version bump now that dansmith is back :)
17:17:27 openstackgerrit Kashyap Chamarthy proposed openstack/nova master: libvirt: Allow disabling CPU flags via `cpu_model_extra_flags` https://review.opendev.org/c/openstack/nova/+/774240
17:17:36 gibi bauzas: awesome :0\
17:17:38 gibi :)
17:17:52 bauzas I was mostly done with it, but faced problems
17:18:23 bauzas now it will require some rebase, as I think we had a new minor bump due to cyborg shelve/unshelve
17:18:40 gibi yepp
17:19:11 bauzas but honestly, the test failure puzzles me
17:22:01 gibi dansmith: hi! as per http://lists.openstack.org/pipermail/openstack-discuss/2021-February/020186.html could you help me adding the nova-stable-maint group to the placement-stable-maint group in gerrit. I'm not in the latter so I cannot do the change
17:22:31 bauzas gibi: I can do it
17:22:49 gibi bauzas: no you can't
17:22:58 gibi you are not in the placement-stable-maint as far as I see
17:23:02 gibi https://review.opendev.org/admin/groups/7b73c52bbb180a693bac4f2fcca01a6dab9c1222,members
17:23:12 bauzas ah shit indeed, the other way :)
17:23:28 bauzas and not a stable-maint-core :)
17:23:52 gibi :)
17:34:29 sean-k-mooney gibi: as far as i know its invalid for a port with multiple fixed_ip from different segments to be created
17:34:58 sean-k-mooney gibi: there may be one excption
17:35:22 sean-k-mooney having ips on differnet segment implies that port is on multiple differnt l2 broadcast domain concurrently
17:35:32 sean-k-mooney which is not physically possible
17:35:54 sean-k-mooney the caveate which i think we should declare out of scope would be a situration where you are using routed network with trunk portrs
17:36:36 sean-k-mooney in principal each subport could be on a different physical network and the the parent port could therefor hoave fixed ips form different segments i think
17:37:15 sean-k-mooney gibi: we might be able to support that at some point but for now i think we would be better off documenting that as a limition in this release
17:37:27 sean-k-mooney an enable that if someone asks for it in the future
17:38:01 sean-k-mooney gibi: i think multiple fixed ips is fine just they should have the same sgement
17:38:32 sean-k-mooney they could have different subnets e.g. if you ran out of ips and wantted to add more to a segment via a second subnet
17:38:49 sean-k-mooney but 1 segment and 1 l2 broadcast domain i hope that makes sense?
17:39:06 sean-k-mooney ill review bauzas patch in either case
17:40:43 sean-k-mooney trunk ports being https://specs.openstack.org/openstack/neutron-specs/specs/newton/vlan-aware-vms.html
17:51:05 openstackgerrit Kashyap Chamarthy proposed openstack/nova master: libvirt: Allow disabling CPU flags via `cpu_model_extra_flags` https://review.opendev.org/c/openstack/nova/+/774240
17:51:33 kashyap gibi (and others): For tomm --^ :)
17:51:58 kashyap gibi: You must be drowning in review requests. Sorry for the deluge
18:04:06 sean-k-mooney https://review.opendev.org/c/openstack/nova/+/749068/12/nova/scheduler/request_filter.py#312 commented inline
18:04:40 dansmith gibi: sorry was on a call, looking now
18:05:20 openstackgerrit Stephen Finucane proposed openstack/nova master: mypy: Add type annotations to 'nova.pci' https://review.opendev.org/c/openstack/nova/+/676209
18:05:21 openstackgerrit Stephen Finucane proposed openstack/nova master: mypy: Add nova.cmd, nova.conf, nova.console https://review.opendev.org/c/openstack/nova/+/705657
18:05:22 openstackgerrit Stephen Finucane proposed openstack/nova master: mypy: Add type annotations to top-level modules https://review.opendev.org/c/openstack/nova/+/705658
18:05:23 openstackgerrit Stephen Finucane proposed openstack/nova master: trivial: Clean manager.Manager, service.Service signatures https://review.opendev.org/c/openstack/nova/+/764806
18:06:12 dansmith gibi: how's that?
18:20:34 ganso Hi Nova folks! Does this look like a legit bug to you? I'm thinking of ways to fix it, either change "openstack usage show" to use created_at instead of launched_at (negative side effects are created instances that were launched) or stop updating the launched at field on resize and unshelved. Thinking about it, a cold migration or a stop/start operation does not update the launched field, why should a resize do? https://bugs.launchpad.net/nova/+bug/1
18:20:34 openstack Launchpad bug 1 in Ubuntu Malaysia LoCo Team "Microsoft has a majority market share" [Critical,In progress] - Assigned to MFauzilkamil Zainuddin (apogee)

Earlier   Later