Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2018-11-09
21:25:58 dansmith isn't that a good reason to pack and get out?
21:26:09 mriedem i've just holed up in my office
21:26:26 mriedem i'll crank out a patch for this bug and be off
21:45:39 openstackgerrit Jack Ding proposed openstack/nova master: Use virt.images.convert_image for qemu-img convert https://review.openstack.org/616692
21:52:35 fried_rice mriedem: Sanity check, please. The compute manager has a report client via the scheduler client, that's *not* the same as the report client the resource tracker has.
21:53:19 fried_rice which means my current SIGHUP doesn't do shit to the RT's cache
21:53:26 fried_rice I need to make the report client a singleton.
21:53:42 mriedem correct
21:53:46 mriedem we have report clients all over the place
21:53:50 mriedem api, conductor, scheduler
21:53:52 mriedem etc
21:54:14 fried_rice that's a scroo
21:54:55 fried_rice mriedem: So - make the report client a singleton (per process), or just diddle the compute manager's reset to hit the rt's reportclient instead.
21:55:59 fried_rice f, without knowing what the various ones in the compute manager are used for, is it really safe to make them a singleton?
21:56:04 mriedem the latter would be a smaller blast area
21:57:42 fried_rice I think I may have actually done this to myself, by removing that LazyLoader
21:57:50 fried_rice I suspect that guy was incidentally singleton-ing.
21:58:33 fried_rice nah, that should still have been creating separate instances per scheduler client.
22:03:56 dansmith fried_rice: yeah I thought that was making it a singleton a month ago when I was looking at it
22:04:18 dansmith a lot of stuff in nova used to be lazy loaded because.. um, terrible reasons
22:04:26 dansmith lazy loaded or pluggable
22:04:52 fried_rice dansmith: In this case it was supposedly because of a circular import. Whether that was ever really an issue, it isn't now, so I ripped it out. But having just looked, I still don't think it was making the report client a singleton. Care to confirm?
22:05:22 dansmith I think I confirmed that a month ago when I was looking into a seemingly recent memory leak
22:05:51 dansmith so I think it's fine that it's gone
22:06:07 dansmith I agree that randomly making it a singleton now should be done with care
22:06:23 dansmith but I don't really know how to convince myself that it's okay once its done, tbh
22:06:45 openstackgerrit Jack Ding proposed openstack/nova master: Use virt.images.convert_image for qemu-img convert https://review.openstack.org/616692
22:07:25 fried_rice well, using the RT's report client fixed the problem I was having. So maybe I pretend singleton was never suggested.
22:10:15 fried_rice oh, f, this is gonna break all over the place. I can't see a reason why the compute manager would possibly want or need to use separate report clients. I'd really like to put 'em together. If not making it a singleton, at least using only one of them from the compute manager.
22:10:46 fried_rice o/
22:15:46 mriedem fried_rice: the compute manager / RT using the same report client is probably fine,
22:16:06 mriedem a lot of that compute manager / RT code was cleaned up way back in ocata i think when jaypipes made the RT a singleton that tracked multiple compute nodes,
22:16:12 mriedem whereas before it was 1 RT per compute node
22:16:12 fried_rice mriedem: Ima put up an independent patch for that
22:16:14 fried_rice ah
22:16:33 mriedem they are very tightly coupled, like how the compute manager passes the virt driver into the RT
22:16:34 openstackgerrit Jack Ding proposed openstack/nova master: Use virt.images.convert_image for qemu-img convert https://review.openstack.org/616692
22:18:25 openstackgerrit Eric Fried proposed openstack/nova master: Reduce calls to placement from _ensure https://review.openstack.org/615677
22:18:25 openstackgerrit Eric Fried proposed openstack/nova master: SIGHUP n-cpu to refresh provider tree cache https://review.openstack.org/615646
22:18:26 openstackgerrit Eric Fried proposed openstack/nova master: Commonize _update code path https://review.openstack.org/615705
22:18:26 openstackgerrit Eric Fried proposed openstack/nova master: Consolidate inventory refresh https://review.openstack.org/615695
22:18:27 openstackgerrit Eric Fried proposed openstack/nova master: Turn off rp association refresh in nova-next https://review.openstack.org/616033
22:18:31 fried_rice let's see how that goes
22:21:44 fried_rice mriedem: Oh, my removal of lazyload probably reinstated "lockutils spam" mentioned in nova/compute/api.py@257
22:25:15 openstackgerrit Matt Riedemann proposed openstack/nova master: Retry on consumer delete race in claim_resources https://review.openstack.org/617040
22:25:17 mriedem dansmith: jaypipes: fried_rice: ^ bingo bango
22:25:36 mriedem gibi: you too ^
22:25:48 mriedem the commit message is longer than the code
22:29:02 mriedem and with that i'm off
22:52:41 openstackgerrit Eric Fried proposed openstack/nova master: Rip the report client out of SchedulerClient https://review.openstack.org/617042
22:54:58 openstackgerrit Eric Fried proposed openstack/nova master: Rip the report client out of SchedulerClient https://review.openstack.org/617042
23:17:20 openstackgerrit Vlad Gusev proposed openstack/nova stable/pike: libvirt: Reduce calls to qemu-img during update_available_resource https://review.openstack.org/604039
23:18:08 openstackgerrit Merged openstack/nova stable/pike: Make scheduler.utils.setup_instance_group query all cells https://review.openstack.org/599841
23:18:46 openstackgerrit Vlad Gusev proposed openstack/nova stable/pike: libvirt: Reduce calls to qemu-img during update_available_resource https://review.openstack.org/604039
23:20:37 openstack bug 1799892 in OpenStack Compute (nova) rocky "Placement API crashes with 500s in Rocky upgrade with downed compute nodes" [Medium,Triaged] https://launchpad.net/bugs/1799892
23:20:37 openstackgerrit Merged openstack/nova master: Add recreate test for bug 1799892 https://review.openstack.org/613304
23:25:52 aspiers mriedem: thanks for the review! Regarding technical debt, my understanding is that the intention is very much for SUSE/AMD to carry on working to flesh out the functionality after implementation of the MVP described in the initial spec, rather than just to dump some half-baked implementation upstream and then vanish ;-) This would include adding support for attestation, migration etc.
23:26:04 aspiers ah, he's gone
23:27:55 openstackgerrit Vlad Gusev proposed openstack/nova stable/pike: libvirt: Use os.stat and os.path.getsize for RAW disk inspection https://review.openstack.org/607544
23:56:50 openstackgerrit Merged openstack/nova master: Mention meta key suffix in tenant isolation with placement docs https://review.openstack.org/616991
23:57:06 openstackgerrit Ken'ichi Ohmichi proposed openstack/nova master: api-ref: Add a description about sort order https://review.openstack.org/616773
#openstack-nova - 2018-11-10
00:56:21 openstackgerrit Eric Fried proposed openstack/nova master: Rip out the SchedulerClient https://review.openstack.org/617049
00:56:21 openstackgerrit Eric Fried proposed openstack/nova master: Rip the report client out of SchedulerClient https://review.openstack.org/617042
01:00:20 openstackgerrit Eric Fried proposed openstack/nova master: Reduce calls to placement from _ensure https://review.openstack.org/615677
01:00:20 openstackgerrit Eric Fried proposed openstack/nova master: SIGHUP n-cpu to refresh provider tree cache https://review.openstack.org/615646
01:00:21 openstackgerrit Eric Fried proposed openstack/nova master: Consolidate inventory refresh https://review.openstack.org/615695
01:00:22 openstackgerrit Eric Fried proposed openstack/nova master: Rip the report client out of SchedulerClient https://review.openstack.org/617042
01:00:22 openstackgerrit Eric Fried proposed openstack/nova master: Commonize _update code path https://review.openstack.org/615705
01:00:23 openstackgerrit Eric Fried proposed openstack/nova master: Rip out the SchedulerClient https://review.openstack.org/617049
01:01:16 openstackgerrit Eric Fried proposed openstack/nova master: Turn off rp association refresh in nova-next https://review.openstack.org/616033
01:30:19 openstackgerrit Merged openstack/nova master: libvirt: change "Ignoring supplied device name" warning to info https://review.openstack.org/616952
01:30:25 openstackgerrit Merged openstack/nova master: Add a link to the doc contrib guide https://review.openstack.org/616063
01:30:32 openstackgerrit Merged openstack/nova master: Fix version details API does not return 200 OK https://review.openstack.org/616444
04:36:46 openstackgerrit Merged openstack/os-vif master: Update min tox version to 2.0 https://review.openstack.org/615046
13:05:05 openstackgerrit Matt Riedemann proposed openstack/nova master: RT: replace _instance_in_resize_state with _is_trackable_migration https://review.openstack.org/560467
13:13:04 openstackgerrit Matt Riedemann proposed openstack/nova master: Use long_rpc_timeout in select_destinations RPC call https://review.openstack.org/607735
13:15:09 openstackgerrit Matt Riedemann proposed openstack/nova master: Fix jsonutils.to_primitive UserWarning https://review.openstack.org/612445
#openstack-nova - 2018-11-12
03:01:56 openstackgerrit JiaJunsu proposed openstack/nova master: Remove args(os=False) in monkey_patch https://review.openstack.org/568999
03:21:39 openstackgerrit zhouxinyong proposed openstack/nova master: modify the avaliable link https://review.openstack.org/616905
04:43:03 openstackgerrit zhouxinyong proposed openstack/nova master: Fix link in get-me-a-network-992eabc81b5e5347 https://review.openstack.org/617222
06:50:28 openstackgerrit Merged openstack/nova master: api-ref: Add a description about sort order https://review.openstack.org/616773
11:30:45 openstackgerrit Lee Yarwood proposed openstack/nova stable/rocky: Harden placement init under wsgi https://review.openstack.org/617297
11:31:26 lyarwood sean-k-mooney: ^ btw, backported your placement hardening patch if you have time this week to ack.
11:38:06 sean-k-mooney thanks that was actully going to be the thing i did next
11:38:35 sean-k-mooney ill review instead
11:50:27 sean-k-mooney lyarwood: how far back should we backport that chage? placement became mandaotry in pike right? i would think once its merged on rocky we would at least want it on queens but i would probaly backport to pike assuming the code is largely unchaged
11:54:05 openstackgerrit Jens Harbott (frickler) proposed openstack/nova master: Make supports_direct_io work on 4096b sector size https://review.openstack.org/616580
11:55:19 lyarwood sean-k-mooney: yeah back to pike would be useful, feel free to post them now if you have time or I can look later today
11:55:54 sean-k-mooney just reading a spec so ill post them shortly.
14:15:14 openstackgerrit Martin Midolesov proposed openstack/nova master: [vmware] Cross vCenter live migration https://review.openstack.org/616590
14:29:23 ThomasWhite Hey all, a quick question I couldn't find a reliable answer to. How often or what would qualify running the "nova-manage db archive_deleted_rows" command? Is this something I should view as a routine operation, or only a as-needed step?
14:29:49 ThomasWhite Our galera cluster is stable and quick, no pressing need for it nor are we short on space, but just looking to implement any best practices
14:33:40 sean-k-mooney ThomasWhite: that is really up to you. if you want to be proactive and keep the current level of perfromace you could proactivly run that command on a fixed cadance
14:34:39 sean-k-mooney ThomasWhite: but if you have monitoring in place to determin you perfromance is within your accpeatable critia then you could decide to only run it when that critia is not met
14:35:46 sean-k-mooney ThomasWhite: if you upgrade every release then running it as part of your upgrade may be suffient if your cloud is relitivly stable with longer lived instnaces
14:36:27 sean-k-mooney ThomasWhite: if you are running a ci environment where you are constantly spawning and deleteing instance then you might want to run it on a cron job every X weeks
14:37:46 sean-k-mooney i personally would only run it on upgrade or when needed as i dont like fixing things that are not broken but perhaps some operators can share there experiance?
14:38:25 ThomasWhite Awesome - thanks sean-k-mooney. Would removal of individual entries from instance_actions do anything bad either - primarily as we need to take a look at a server (booting through rescue) but not leave the regular user of it with knowledge we've taken a look for compliance.
14:39:11 ThomasWhite Putting into single user could impact data reliability so booting into rescue is the preferred mechanism
14:40:14 sean-k-mooney ThomasWhite: that i dont know the anwser too. it if your billing system cannot elimitate teh run time of the insance using o

Earlier   Later