Earlier  
Posted Nick Remark
#openstack-nova - 2021-10-22
13:12:21 gibi sean-k-mooney: fyi, sriov agent only unrealiably sends vif plug for VFs, as it polls the hypervisor. If the unplug/plug is fast enough then the agent might miss the state when the device was down
13:12:42 sean-k-mooney gibi: more fun
13:12:46 sean-k-mooney ok
13:12:52 gibi it is fun all the way down :D
13:13:04 sean-k-mooney so we might want to only wait for vnic_type=normal then
13:13:15 gibi it seems soo
13:13:53 sean-k-mooney long term we really do need to fix this interface and enforce a stricter contract
13:14:05 sean-k-mooney rather then guessing cause there are so many factors to consider
13:14:58 gibi yes, we need neturon to be either enforce that events always sent, or declare in the port what event can be expected. There is no way nova can maintain a sane mapping alone
13:25:58 bauzas dmitriis: saw the chat ?
13:26:13 dmitriis bauzas: looking, 1 sec
13:26:21 bauzas dmitriis: I'm about to propose to postpone your topic after 3pm UTC
13:27:16 dmitriis bauzas: got a conflict at 3PM but I can make it work
13:27:26 bauzas dmitriis: maybe later then ?
13:27:43 bauzas the idea is just to avoid discussing your stuff *before* :)
13:28:11 dmitriis bauzas: let's do it at 3PM, later is more complicated :^)
13:28:27 bauzas ack, moving your topic then :)
13:28:52 dmitriis bauzas: ack, ty for pinging
13:33:57 stephenfin Is it just me or is tbarron's sound clipping real bad? I can understand him though (i.e. can be fixed later)
13:38:28 tbarron stephenfin: It may be from my end, sorry. Yesterday zoom was having trouble with my rural location.
13:39:01 stephenfin tbarron: nw, I was just concerned something was broken on my end :) I could understand everything just fine
13:39:19 tbarron cool
13:43:09 opendevreview Stephen Finucane proposed openstack/nova master: db: Remove models that were moved to the API database https://review.opendev.org/c/openstack/nova/+/812149
13:43:10 opendevreview Stephen Finucane proposed openstack/nova master: db: Remove models for removed services, features https://review.opendev.org/c/openstack/nova/+/812150
13:43:10 opendevreview Stephen Finucane proposed openstack/nova master: db: Remove nova-network models https://review.opendev.org/c/openstack/nova/+/812151
13:45:59 opendevreview Balazs Gibizer proposed openstack/nova stable/pike: Add a WA flag waiting for vif-plugged event during reboot https://review.opendev.org/c/openstack/nova/+/813437
13:56:59 kashyap clarkb: I'm innundated with a few things; I will reply to your email on the list on Monday. Hope that's okay
15:15:59 clarkb kashyap: yup no worries
15:16:02 clarkb have a good weekend
17:04:44 bauzas dansmith: dunno if you're fancy happy joining us, but we're discussing a topic your knowledge could be helpful : instance v3.0 object bump
17:04:58 dansmith sorry, I'm tied up
17:05:01 bauzas or you're stuck in the TC meeting
17:05:06 bauzas heh, no worries
17:05:15 dansmith I definitely have opinions
17:05:18 bauzas dansmith: notes on the etherpad could be appreciated tho :)
17:05:28 bauzas https://etherpad.opendev.org/p/nova-yoga-ptg L646
17:26:43 sean-k-mooney https://etherpad.opendev.org/p/nova-yoga-ptg-backup
17:26:54 sean-k-mooney no colors but there is a snapshot ^
17:28:04 clarkb I just restored it to the version that bauzas identified as good (the original etherpad url I mean)
17:28:22 bauzas yeah \o/
17:28:24 sean-k-mooney yep that looks ok
17:28:39 bauzas I just wrote our last conclusions
17:28:43 bauzas we can call it a wrap
17:28:51 sean-k-mooney clarkb++ thanks
17:28:57 bauzas sean-k-mooney: just copy it again, so we have a backup
17:29:22 sean-k-mooney i just did an plain text export form the timeline and imported it again in adifferent page
17:29:31 bauzas yeah that works
17:29:41 sean-k-mooney but ill try that in html form and see if it can keep colours
17:29:44 bauzas we have the highlights with the history
17:29:57 bauzas so nothing is technically lost
17:30:00 sean-k-mooney in etherpad format it goes to the latest point in history not the one you have selected in the timeline
17:31:07 sean-k-mooney ya so html format does not keep colors either
17:31:18 sean-k-mooney but we have teh backup in any case
17:31:28 sean-k-mooney and the orginal is restored so we are good
17:32:24 bauzas yup
17:32:33 bauzas on that note, /me calls it a week
17:32:36 bauzas \o
17:32:50 gibi me too
17:32:52 gibi o/
17:33:00 bauzas I'm just sad to hear that the next PTG will still be virtual, but that's life
17:33:27 bauzas I'm just exhausted and I miss our whiteboards and hallway discussions
17:34:00 bauzas but I'll open a beer to enjoy the last day of the PTG as if it was physical :)
19:22:37 mnaser_ hm
19:22:52 mnaser_ anyone would have an idea as to why metadata service seems to be stalling / very slow
19:23:25 mnaser_ 2021-10-22 19:23:13.422 11 INFO nova.metadata.wsgi.server [req-f65469d1-4082-4b9e-ab3b-3c710c1f0366 - - - - -] 10.30.107.187,10.101.2.141 "GET /latest/meta-data/block-device-mapping/root HTTP/1.1" status: 200 len: 148 time: 25.8355188
19:23:43 mnaser_ it almost feels like all 'requests' stall for a bit and then they all burst out at once
19:23:47 mnaser_ memcache is fine, mysql is fine
19:24:06 mnaser_ plenty of conductors
19:25:30 mnaser_ 2021-10-22 19:25:15.880 13 INFO nova.metadata.wsgi.server [-] 192.168.1.5,10.101.2.142 "GET /openstack HTTP/1.1" status: 200 len: 235 time: 5.0295317
19:25:35 mnaser_ especailly this, it seems very 5s-y
19:46:53 clarkb mnaser_: I've noticed that on a couple of devstack jobs recently too but haven't had a chance to dig into it beyond noting that was the error returned by tempest
19:47:13 clarkb specifically some tests says metadata service fails to respond in time and the test times out
20:44:05 prometheanfire nova fails with the new oslo-concurrency 4.5.0 https://zuul.opendev.org/t/openstack/build/9a385f3324fb46a3abf6257a09020d38
20:45:07 prometheanfire looks like an extra value was used (blocking)?
#openstack-nova - 2021-10-25
07:22:03 opendevreview Wenping Song proposed openstack/osc-placement master: Remove usage of six https://review.opendev.org/c/openstack/osc-placement/+/815223
07:26:41 opendevreview Wenping Song proposed openstack/osc-placement master: Remove usage of six https://review.opendev.org/c/openstack/osc-placement/+/815223
08:08:03 opendevreview Wenping Song proposed openstack/nova master: Cleanup guest process when vm evacuate failed and then deleted https://review.opendev.org/c/openstack/nova/+/814746
08:23:23 bauzas good morning Nova
08:25:40 gibi bauzas: o/
13:00:19 artom_ johnthetubaguy[m], TheJulia, so here's an idea (to follow up on the node reassignment hash ring changed thing)
13:00:27 artom_ Move the whole thing to the conductor
13:02:16 artom_ We would in effect be moving virt-driver-specific code into the conductor, but it sure would make everything else easier...
13:51:51 kashyap clarkb: Very cool; this is merged - https://review.opendev.org/c/openstack/devstack/+/815020 :)
15:18:05 opendevreview Erlon R. Cruz proposed openstack/nova master: Fix pre_live_migration rollback https://review.opendev.org/c/openstack/nova/+/815324
15:18:45 opendevreview Erlon R. Cruz proposed openstack/nova master: Fix pre_live_migration rollback https://review.opendev.org/c/openstack/nova/+/815324
15:37:51 kashyap clarkb: Responded on the list. Hope that answers your questions. (Even if it's not fully satisfactory)
15:37:58 clarkb kashyap: it does, thanks
15:38:23 clarkb kashyap: also https://bugzilla.redhat.com/show_bug.cgi?id=2010058 seems to be at least part of our fedora problems
16:54:38 opendevreview Merged openstack/nova master: Reproducer unit test for bug 1934094 https://review.opendev.org/c/openstack/nova/+/812646
17:42:06 gibi melwitt, sean-k-mooney[m]: I think https://review.opendev.org/c/openstack/nova/+/814036 is the way to go. My alternative trial to kill leaking eventlet failed in https://review.opendev.org/c/openstack/nova/+/815017
18:27:11 TheJulia artom: hmm, I don't know the nova conductor code path well enough to really know for sure. That feels like we're just shifting a point of failure, but at the same time the compute hosts are basically stateless and just fulfill a role...
18:30:17 artom TheJulia, it might be workable if we build a decent abstraction around it... like, the conductor gets the list of running compute services, calls them and passes the list, the services+driver combo do the hash ring thing and return the result to the conductor, the conductor the saves that
18:30:29 artom Really rough initial first idea
18:33:09 TheJulia artom: I guess the question I would ahve is what spured the thought to think to move it
18:35:48 artom TheJulia, well the current problem is that in your proposed patch we'd be modifying instance.host and potentially doing it while something else is accessing it. Except we can't use the lock that everything else uses because we have no guarantee that we're on the same compute host
18:35:58 artom And the lock is host-local
18:37:12 TheJulia Well, it is a single field update on a row, the db should lock it for the transaction unless sqlachemy's save is update all fields or insert/delete
18:37:29 TheJulia and *if* something should happen where there is a split perception, it should reconcile
18:37:34 TheJulia on the net run
18:37:49 TheJulia s/net/next

Earlier   Later