| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-08-29 | |||
| 12:28:53 | stephenfin | alex_xu: If you'd like, I can split that change out into its own fix so we can maybe backport it (though it's low priority) | |
| 12:29:14 | stephenfin | I think we should have a release note too, just in case | |
| 12:29:27 | alex_xu | stephenfin: I'm ok with it now, since agree with that it is corner case. maybe release note is enough | |
| 12:29:28 | stephenfin | I'll tack both on at the end and move them in if/when I need to rebase | |
| 12:29:38 | stephenfin | cool, reno it is | |
| 12:29:41 | alex_xu | cool | |
| 12:30:13 | alex_xu | I also didn't found any problem with https://review.opendev.org/#/c/671800/22, I think what i need is test now | |
| 12:30:57 | stephenfin | alex_xu: There are quite a few functional tests but they're at the end of the series https://review.opendev.org/#/c/671801 | |
| 12:31:26 | alex_xu | cool, I need to check it out | |
| 12:31:30 | stephenfin | I think sean-k-mooney is planning to test manually too. I was hoping to do the same myself but I've been prioritizing reviews | |
| 12:32:12 | stephenfin | Oh, and Bhagyashri Shewale (NTT) has been testing it for me. I should get them to leave comments if it works as expected | |
| 12:32:21 | alex_xu | stephenfin: the first test I want to do is resize | |
| 12:32:35 | stephenfin | Oh, I bet I don't have one for that | |
| 12:32:51 | stephenfin | I can put that together later this afternoon though | |
| 12:32:52 | alex_xu | \o/ | |
| 12:34:04 | sean-k-mooney | stephenfin: test which PCPUs in placment | |
| 12:34:07 | sean-k-mooney | or something else | |
| 12:34:12 | stephenfin | yarp | |
| 12:34:17 | sean-k-mooney | ah yes | |
| 12:34:39 | sean-k-mooney | i plan to test that preferably with artroms and gibis stuff too | |
| 12:34:57 | sean-k-mooney | so hopfully some of this will start merging to make that simpelre to do | |
| 12:35:18 | sean-k-mooney | when you think its a good time to test it let me know and ill test it standalone | |
| 12:36:49 | stephenfin | sean-k-mooney: I think it's good to go now, tbh | |
| 12:37:55 | sean-k-mooney | ok are there specific things you would like me to test | |
| 12:38:10 | stephenfin | Yup, if you give me five I can put together a list of test cases for things that I think should be covered | |
| 12:38:37 | sean-k-mooney | cool add it to an etherpad | |
| 12:39:00 | sean-k-mooney | i can then run through it add updated it with my results | |
| 12:39:04 | alex_xu | I guess resize is just won't be trigger between old and new, since there is no PCPU on old node. | |
| 12:39:44 | stephenfin | Yeah, a resize from a host reporting PCPU to one reporting VCPU won't be possible since it won't get past the scheduler | |
| 12:39:56 | sean-k-mooney | well that depends | |
| 12:40:08 | sean-k-mooney | if you resize form pinned to non pinned it woudl be | |
| 12:40:15 | stephenfin | Good point, actually | |
| 12:40:36 | sean-k-mooney | a cold migrate would not be | |
| 12:40:38 | alex_xu | oh | |
| 12:41:30 | stephenfin | We have functional coverage for cold migrate (for the NUMA-aware vSwitches feature) already but not resize. Should be easy enough add | |
| 12:41:58 | sean-k-mooney | ya its the same test just with two flavors instead of 1 | |
| 12:42:17 | sean-k-mooney | well kind of | |
| 12:45:32 | alex_xu | anyway, i want to end today early. see you guys tomorrow~ | |
| 12:45:45 | stephenfin | o/ | |
| 12:53:25 | openstackgerrit | Merged openstack/nova stable/stein: Delete unused get_all_host_states method https://review.opendev.org/676255 | |
| 12:53:37 | stephenfin | sean-k-mooney: https://etherpad.openstack.org/p/nova-cpu-resources | |
| 13:01:00 | sean-k-mooney | cool ill try and find time to do that proably on tuesday but ill see if i can get to it sooner. | |
| 13:01:21 | sean-k-mooney | stephenfin: do you want me to test on real hardware or vms | |
| 13:01:31 | stephenfin | VMs should be fine, right? | |
| 13:01:40 | stephenfin | there's nothing hardware specific about this | |
| 13:02:05 | sean-k-mooney | ok if i dont need to use the hardware im using to test artoms stuff that makes it simpler to test | |
| 13:15:23 | openstackgerrit | Merged openstack/nova stable/stein: Convert HostMapping.cells to a dict https://review.opendev.org/676256 | |
| 13:15:30 | openstackgerrit | Merged openstack/nova stable/stein: Cache host to cell mapping in HostManager https://review.opendev.org/676257 | |
| 13:15:38 | openstackgerrit | Merged openstack/nova master: Revert "Filter UnsupportedServiceVersion warning" https://review.opendev.org/677825 | |
| 13:17:52 | openstackgerrit | Kobi Samoray proposed openstack/nova master: Avoid fetching metadata when no subnets found https://review.opendev.org/679247 | |
| 13:21:27 | efried | stephenfin: you around? I keep trying to context-switch in before asking you these questions, but it's always too late by then. So I'm going to be a bit off the cuff here... | |
| 13:21:38 | stephenfin | I am | |
| 13:21:40 | stephenfin | go for it | |
| 13:21:49 | openstackgerrit | Shilpa Devharakar proposed openstack/nova master: DB API changes to get non-matching aggregates from metadata https://review.opendev.org/671074 | |
| 13:21:49 | openstackgerrit | Shilpa Devharakar proposed openstack/nova master: Nova object changes for forbidden aggregates request filter https://review.opendev.org/671072 | |
| 13:21:50 | openstackgerrit | Shilpa Devharakar proposed openstack/nova master: Enable request filter isolate_aggregates https://review.opendev.org/667952 | |
| 13:21:50 | openstackgerrit | Shilpa Devharakar proposed openstack/nova master: Add a new request filter to isolate aggregates https://review.opendev.org/671075 | |
| 13:23:03 | efried | stephenfin: okay: Am I crazy or do I remember somewhere in the cpu-resources series where you are making a design decision based on the assumption that operators are supposed to be segregating their dedicated instances from their shared instances? | |
| 13:23:21 | stephenfin | that was there but it's gone now | |
| 13:24:23 | efried | okay, cause my next question was going to be: ...but post- the feature, you *can* put shared and dedicated instances on the same host, right? | |
| 13:24:33 | efried | ...just not shared & dedicated on the same *guest*. | |
| 13:24:33 | stephenfin | correct | |
| 13:24:37 | stephenfin | yup | |
| 13:24:58 | efried | cool. Cause otherwise... not much point configuring both shared & dedicated set, eh? :P | |
| 13:25:02 | efried | okay, next thing... | |
| 13:26:20 | efried | (in -oslo) | |
| 13:30:21 | mriedem | dansmith: i've updated this prep resize at dest cross-cell patch to remove some of the unused stuff and cleanup the commit/docstrings https://review.opendev.org/#/c/633293/46 - i'm still not really sold on munging this into the existing prep_resize and tried to explain why in the commit and docstring on the method if that helps | |
| 13:31:18 | mriedem | i of course can if it's going to be a blocker, i think it would mean changing prep_resize to consider if migration.cross_cell_move is True and if so, don't reschedule and don't cast to resize_instance on the source | |
| 13:34:23 | bauzas | efried: I think we stated a lot of implications between shared and pinned instances in the spec | |
| 13:35:58 | efried | bauzas: Yeah, I just got a poke from elsewhere that made me remember seeing that assumption/limitation in the code somewhere and I was like "wha...?" | |
| 13:36:20 | bauzas | efried: I'd recommend you to open the spec while you review the cpu-resources series | |
| 13:36:29 | bauzas | that and take some aspirin | |
| 13:36:47 | bauzas | I've started reviewing it but I'm not yet able to sharpen my pen | |
| 13:36:50 | efried | yes, I have it (the in-flight edited version) open. | |
| 13:37:24 | efried | I've been thrashing like a 386 running minecraft though, been reviewing like three patches every other day. | |
| 13:42:03 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Move live_migration test hooks under gate/ https://review.opendev.org/679251 | |
| 13:44:28 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Rename the nova-grenade-live-migration job to nova-grenade-multinode https://review.opendev.org/678841 | |
| 13:44:40 | openstackgerrit | Merged openstack/nova master: Docs and functional test for max_local_block_devices https://review.opendev.org/678268 | |
| 13:44:56 | openstackgerrit | Merged openstack/nova master: Add a "Caveats" section to the eventlet profiling docs https://review.opendev.org/676672 | |
| 13:45:05 | openstackgerrit | Merged openstack/nova master: docs: pretty up return code table or heal_allocations https://review.opendev.org/678063 | |
| 13:45:18 | openstackgerrit | Merged openstack/nova master: doc: pretty up return code table for sync_aggregates https://review.opendev.org/678064 | |
| 13:46:59 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Add cold migrate and resize to nova-grenade-multinode https://review.opendev.org/679210 | |
| 13:47:38 | Sundar | mriedem: Re. the comment in https://etherpad.openstack.org/p/nova-runways-train that tests are failing on Cyborg patches, as of now, they are few and seem unrelated to the code. Even back then, there was really only 1 failure related to the patches. What can we do to get reviews on these patches? | |
| 13:49:48 | openstackgerrit | Merged openstack/nova master: rbd: use MAX_AVAIL stat for reporting bytes available https://review.opendev.org/556692 | |
| 13:52:09 | openstackgerrit | Merged openstack/nova master: Switch some GitHub URLs to point to opendev.org https://review.opendev.org/677185 | |
| 13:52:17 | openstackgerrit | Merged openstack/nova stable/stein: Add useful error log when _determine_version_cap raises DBNotAllowed https://review.opendev.org/675714 | |
| 13:53:11 | mriedem | Sundar: fair enough, i just looked through the series and the failed tests and indeed they look unrelated, just need rechecking | |
| 13:53:30 | mriedem | as for review, i didn't know it was ready yet - i thought most of the work was being done on the cyborg side (i see the series there isn't merged yet either) | |
| 13:53:54 | donnyd | efried: so it looks like changing consecutive_build_service_disable_threshold=0 solved the issues with ghost disabling of the FN hypervisors. Now I just need to find the proper weight to associate with build_failure_weight_multiplier | |
| 13:54:19 | mriedem | Sundar: it might be good to post something short to the mailing list given a status update, e.g. where the nova patches are (what's ready, what's not - known limitations), and what's going on on the cyborg side - and what the status is on getting a fake cyborg driver for integration testing with tempest going so we can have more confidence in the nova patches | |
| 13:55:07 | mriedem | donnyd: i don't think you do if you set consecutive_build_service_disable_threshold=0 | |
| 13:55:17 | mriedem | 0 * any weight multiplier is 0 | |
| 13:55:47 | donnyd | mriedem: I am going to set that back to an appropriate value, was just making sure that was the reason the hypervisors were being ghost disabled | |
| 13:56:31 | donnyd | I don't want actually broken hypervisors to be in system if nova will take care of that for me | |
| 13:57:19 | donnyd | But i am thinking changing it back to something reasonable, and then setting a weight that doesn't take it out of system for too long would be a good resolution | |
| 13:57:54 | donnyd | but I also may not fully understand how build_failure_weight_multiplier works... so I am going to science the crap out of it | |
| 13:58:45 | mriedem | ♪ she blinded me with science ♪ | |
| 14:02:52 | Sundar | mriedem: Yes, but we are able to test with the Cyborg patches. We will merge them soon. | |
| 14:03:00 | artom | Huh, looks like the NUMA functional tests never actually used the NUMATopologyFilter... | |
| 14:03:02 | bauzas | stephenfin: I'm pretty done with the first patch but I'm confused about one last bit | |