Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-03-31
16:16:40 gibi OK, then I suggest to do the reordeing. If swap fails then we need to accept a partial PUT
16:16:47 gibi as swap is async
16:16:51 dansmith ack
16:16:59 dansmith gibi: by the way, on your comment about the notification test,
16:17:19 dansmith when enabling debug, I did not see the compute manager code logging the messages that should have come out
16:17:44 dansmith so I'm not sure the compute manager code is really running, even though that *seems* to be where the finish notification comes from
16:17:50 gibi with 2.85 you don't as the whole test failed before the compute had a chance to progress with the message
16:18:05 dansmith but I ran out of time trying to trace that down and the logs weren't coming..
16:18:14 dansmith ahh, that makes sense
16:18:35 gibi yeah, it was not a trivial thing. I spent at least an hour figuring it out why it behaves differently
16:18:56 dansmith the notification stuff is too confusing in general so I didn't trust myself
16:19:13 dansmith and with no logs from compute manager... I tried putting castascall on there, butI got many more errors
16:19:50 dansmith gibi: can you check my reply here? https://review.opendev.org/#/c/693828/22/nova/api/openstack/compute/schemas/volumes.py
16:19:55 gibi looking
16:21:14 gibi dansmith: you copy the POST req schema, but that is different from the schema of the GET response
16:21:47 gibi if we want that the client GET the current attachment, change a field in it, then PUT it back, then PUT needs to accept the format of the GET response
16:22:02 gibi or did I missunderstood the intention here?
16:22:21 dansmith let me go look for an example of a get
16:23:09 dansmith are you saying because volumeAttachments[] vs volumeAttachment ?
16:23:59 dansmith 2.79 GET of a single attachment looks the same to me, no?
16:24:45 gibi it has id in it
16:24:49 dansmith https://docs.openstack.org/api-ref/compute/?expanded=update-server-detail,list-volume-attachments-for-an-instance-detail,show-a-detail-of-a-volume-attachment-detail#show-a-detail-of-a-volume-attachment
16:25:01 tobias-urdin doing upgrade testing after upgrade nova-compute throws a RuntimeError: maximum recursion depth exceeded while calling a Python object http://paste.openstack.org/show/791412/
16:25:18 gibi crate only allows volumeId and device
16:25:21 gibi create
16:25:26 dansmith ah, okay gotcha
16:25:26 tobias-urdin based on traceback i've tried upgrading oslo.db 5.0.2 and oslo.concurrency 3.30.0 (and all other oslo for that matter) but it won't start
16:25:38 gibi you added tag and d-o-t
16:25:38 dansmith I thought you were saying it was structurally quite different or something
16:25:50 gibi I meant we have different amout of fields
16:26:18 gibi structure seems to be the same
16:26:32 dansmith okay I guess I missed serverId although I thought I had it in there
16:26:36 gibi just GET returns extra fields which PUT does not allow
16:26:47 dansmith yep, I see now
16:27:03 gibi serverId and id (which is the attachment id based on the doc)
16:31:51 gmann dansmith: gibi : should we allow tag update also in this - https://review.opendev.org/#/c/693828/21/nova/api/openstack/compute/volumes.py@475
16:32:46 dansmith gmann: sorry I should have said something, but I looked into it and it requires an rpc call to update that, so I think we should punt
16:33:51 gmann dansmith: ohk. it's not our DB only things. got it.
16:33:56 dansmith yeah
16:35:50 tobias-urdin anybody has an idea what could be causing the traceback? :) http://paste.openstack.org/show/791412/
17:07:22 openstackgerrit Stephen Finucane proposed openstack/nova master: Remove future imports https://review.opendev.org/714675
17:07:22 openstackgerrit Stephen Finucane proposed openstack/nova master: Use unittest.mock instead of third party mock https://review.opendev.org/714676
17:26:39 openstackgerrit Merged openstack/os-vif master: Update hacking for Python3 https://review.opendev.org/715651
17:44:30 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Introduce scope_types in lock server policy https://review.opendev.org/716114
17:47:25 sean-k-mooney stephenfin: didnt efried already sumbit a patch for https://review.opendev.org/#/c/714676/
17:47:38 sean-k-mooney did he abandon it?
17:49:11 sean-k-mooney stephenfin: https://review.opendev.org/#/c/708262/
17:54:01 sean-k-mooney stephenfin: it looks like you have fixed import ordering and some other minor thinks so i guess your patch is more complete but you should probably cherry-pick the hacking change ontop of your patch
18:00:14 openstackgerrit Dan Smith proposed openstack/nova master: Allow PUT volume attachments API to modify delete_on_termination https://review.opendev.org/693828
18:01:01 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Add new default roles in lock server policies https://review.opendev.org/716122
18:33:18 efried stephenfin, sean-k-mooney: feel free to take over or abandon mine as needed.
18:33:44 efried but yeah, the hacking change should be included in the series.
19:01:11 openstackgerrit Corey Bryant proposed openstack/nova stable/queens: hardware: fix memory check usage for small/large pages https://review.opendev.org/716328
19:06:22 openstackgerrit Marcin Juszkiewicz proposed openstack/nova master: libvirt: check for AMD SEV only on x86-64 https://review.opendev.org/714425
19:10:00 stephenfin sean-k-mooney: Ah, I'd forgotten about that. Will combine
19:33:47 openstackgerrit Merged openstack/nova master: [Community goal] Update contributor documentation https://review.opendev.org/712420
19:33:57 openstackgerrit Merged openstack/nova master: Fix os-ips policy to be admin_or_owner https://review.opendev.org/715496
19:34:04 openstackgerrit Merged openstack/nova master: Add test coverage of existing ips policies https://review.opendev.org/715477
19:34:11 openstackgerrit Merged openstack/nova master: Introduce scope_types in os-ips https://review.opendev.org/715529
19:34:18 openstackgerrit Merged openstack/nova master: Add new default roles in os-ips policies https://review.opendev.org/715545
20:03:23 openstackgerrit Merged openstack/nova master: Pass the actual target in os-agents policy https://review.opendev.org/701649
20:46:39 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Pass the actual target in unlock override policy https://review.opendev.org/716428
20:55:01 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Pass the actual target in limits policy https://review.opendev.org/715761
20:55:06 dansmith gmann: you around?
20:55:21 dansmith owalsh has a question for you :)
20:55:37 gmann dansmith: hi
20:55:51 dansmith gmann: http://paste.openstack.org/show/kckeKCDgq679ixa2g5IK/
20:56:03 dansmith gmann: when running nova-manage with an empty policies file
20:56:44 owalsh nova-manage cell_v2 discover_hosts --by-service if it matters
20:58:22 dansmith owalsh: I gotta step away for a bit to get ready for something, but from poking around I feel like gmann is probably your mann :)
20:58:33 owalsh dansmith: ack, thanks
20:58:34 gmann :)
20:58:40 openstackgerrit sean mooney proposed openstack/nova master: [WIP] cyborg evacuate support https://review.opendev.org/715326
20:59:31 owalsh gmann: already covered by one of the bp/policy-default-refresh patches?
20:59:38 gmann owalsh: those are new defaults and it add the deprecation even you have not override it. but old token keep working as they are maintained as deprecated rule till now.
20:59:46 gmann yeah
21:00:38 owalsh gmann: cool, so should go away once they all merge
21:00:53 gmann owalsh: we are changing all the policy to adopt the system scope and new defaults like read-only etc
21:03:36 gmann owalsh: warnings will stay till we remove the deprecated old rules in 1 or 2 cycle from now
21:05:07 owalsh gmann: ack, thanks, expect I'll get asked about the warning quite often until then :-)
21:09:23 gmann owalsh: basically signal to adopt the scope check, those are configurable for now and disabled as default. our goal is to 1. remove the old deprecated roles 2. enable scope check together. but you can always move to new policy by configuring the enforce_scope=Ture
21:09:59 gmann i am still working on those and should provide a doc on 'how to migrate to new policy' once done.
21:10:38 gmann owalsh: i hope those warnings are not much disturbing (as they are for every rule).
21:11:12 melwitt gmann: ++ I was just gonna say I wonder if we can improve that warning message and link to a doc explaining the details and process for migrating. a main point in it is the user being able to tell whether they need to migrate at all, based on their existing policy
21:11:43 melwitt at a minimum once you have written a doc, we should update the warning message to include the link to the documentation
21:12:42 gmann melwitt: that is good idea. it will be easy to link doc to warning
21:21:28 owalsh gmann: so for default (empty) policy file we will still get warning for 1-2 cycles?
21:21:49 owalsh or just if deprecated roles are used in the policy?
21:22:03 dansmith it definitely sucks to warn about deprecated policy when the deprecated things are defaults
21:22:21 dansmith if we can't tell what is deprecated (defaults vs. overrides) I would argue we should squelch that warning
21:25:52 gmann owalsh: all becasue defaults are deprecated
21:27:08 gmann it is for all rule as we use those defaults rules as check_str for every rule and oslo policy just add warning for those
21:27:25 gmann not sure how to combine those.
21:30:43 gmann one way is disable oslo warning completely and add a single combined warning form nova policy code with link to migration doc.
21:34:03 bnemec The deprecations are warning you that something might break next cycle and you should test with the future defaults now.
21:35:50 gmann bnemec: we test with both old and new but with disable wanring.
21:36:07 dansmith bnemec: sounds like he has deprecated our existing defaults
21:36:13 dansmith which is why I think we should *not* show this to the user
21:36:20 dansmith there's nothing they can do about it, other than ignore
21:36:46 dansmith it teaches our users to ignore deprecation warnings which is majorly uncool

Earlier   Later