| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2023-04-25 | |||
| 21:25:30 | dansmith | well, either way.. if we move this to only >=jammy for everything (along with the PTI for 2023.2) then we can just use this debs list thing and remove the focal-specific install stuff.. whatever you ant | |
| 21:25:33 | dansmith | I just want it to work :) | |
| 21:25:50 | gouthamr | agree; lets see this work | |
| 21:27:14 | gouthamr | if the "rbd remove" thing fails again on this run with the error, i would suggest reporting a bug - we could _try_ this thing on centos-9-stream and see whether there's some weirdness in the distro packages | |
| 21:27:58 | gouthamr | but, i am nervous about that job's future with ubuntu since we've learned about the ceph community's stance | |
| 21:28:03 | dansmith | ack, well, based on how this is working, I'm hoping the cephadm will either "work" or "fail the same way" and then we can discount distro packages | |
| 21:28:11 | gouthamr | ++ | |
| 21:28:27 | dansmith | it's running tempest now and looking identical to the distro version so far (i.e. hasn't failed but hasn't run volume tests yet) | |
| 21:28:46 | dansmith | so that's majorly better. failing with upstream bits is a minor win over failing with distro bits :P | |
| 21:29:03 | eharney | i have some work in flight currently around rbd ImageBusy errors... i wonder if this job is running some tests that were previously disabled? | |
| 21:29:31 | dansmith | eharney: shouldn't be, and one run failed with a bunch of them and then a single recheck failed similarly, but with no rbdbusy errors | |
| 21:31:40 | gouthamr | eharney: interesting, are there librbd changes that you're having to work around? | |
| 21:32:15 | eharney | gouthamr: not new ones, just working on fixing the class of errors around rbd images that can't be deleted that we've always had | |
| 21:34:57 | gouthamr | eharney: oh.. this error seems to have occurred multiple times in the libvirt rbd "remove image" call.. | |
| 21:36:04 | gouthamr | i'm hoping no openstack code needs to change; we're hoping to support quincy with stable/wallaby downstream :D based on some testing of this stuff elsewhere | |
| 21:36:36 | dansmith | gouthamr: yeah I was going to ask earlier.. are we using quincy downstream such that we know it works? this set of failures has me concerned that it's something fundamental of course | |
| 21:37:14 | gouthamr | (same tests, different OS, full blown ceph cluster as opposed to our aio .. etc etc -- so there could be a number of things being issues) | |
| 21:38:41 | gouthamr | dansmith: yes, we're not testing quincy with openstack's trunk though... we trail downstream; but this stuff is working with zed last i checked, with the same tempest tests passing there | |
| 21:38:53 | dansmith | okay that's good | |
| 21:56:15 | dansmith | well, it's passing some volume tests at least | |
| 21:56:29 | gouthamr | ++ | |
| 21:57:22 | gouthamr | a couple of things we wanted to try in this cephadm job in the past: | |
| 21:57:44 | dansmith | gouthamr: so if this works magically, you're okay just making this drop support for focal as long as the other jobs here are set to run on jammy? | |
| 21:57:46 | gouthamr | (1) revert to default test concurrency --- the concurrency was set to 1 because we saw resource contention | |
| 21:58:00 | gouthamr | dansmith: yes | |
| 21:58:02 | dansmith | resource contention like memory/ | |
| 21:58:10 | gouthamr | yes, and disk | |
| 21:58:26 | dansmith | so there's a tweak in devstack for memory that has been helping a lot of jobs, and we run with it enabled in the nova ceph jobs | |
| 21:58:39 | gouthamr | oh? i'd love to know! | |
| 21:58:43 | dansmith | drops mysql usage by about half, which is ~400MiB on these jobs | |
| 21:58:47 | gouthamr | nice | |
| 21:58:55 | dansmith | gouthamr: problem is you have to pay me a royalty per job execution to use it | |
| 21:59:17 | dansmith | gouthamr: https://github.com/openstack/nova/blob/master/.zuul.yaml#L422 | |
| 21:59:33 | gouthamr | :D i'd be poor if i was betting on ceph using less memory ever | |
| 21:59:38 | dansmith | haha | |
| 22:00:03 | gouthamr | (2) there's also a way to turn off "cephadm" after the install -- we should set that option on the CI - there's no use to keep it running | |
| 22:00:06 | dansmith | so based on our success with that, if you're cool with it, I'd say we turn that on for these jobs anyway | |
| 22:00:24 | dansmith | gouthamr: meaning before we start running tempest? | |
| 22:01:30 | gouthamr | yes, the plugin will do that for you after the ceph cluster deployment is done: https://github.com/openstack/devstack-plugin-ceph/blob/563cb5deeb21815ce0c62fa30249e85e886c783a/devstack/lib/cephadm#L28 | |
| 22:02:12 | dansmith | oh so why is that not being set? | |
| 22:02:33 | dansmith | I don't even know what that means.. cephadm is a tool I thought, but you're saying it stays running even though the services are started or something? | |
| 22:02:36 | gouthamr | (here's an example from the only good cephadm job at the moment: https://github.com/openstack/manila-tempest-plugin/blob/ad0db6359c8c51c4521ac6660c8014981b2f1dea/zuul.d/manila-tempest-jobs.yaml#L413) | |
| 22:02:49 | dansmith | ack | |
| 22:02:56 | gouthamr | yes; we need it for day2 operations on the cluster | |
| 22:03:07 | gouthamr | so if you're running this on your local devstack, it's useful | |
| 22:03:13 | dansmith | ah, sure, okay | |
| 22:03:24 | dansmith | we should set that in these jobs that everyone inherits from | |
| 22:03:36 | gouthamr | ++ | |
| 22:03:42 | dansmith | so maybe a follow-on to this to set that, the mysql thing, and make this job voting | |
| 22:04:00 | dansmith | you know, if and when it starts working :) | |
| 22:04:40 | gouthamr | ^^ +1 | |
| 22:06:02 | dansmith | hmm, gouthamr I just noticed that we're still set to release=pacific for the distro-package-based job | |
| 22:06:11 | dansmith | is that maybe setting stuff that is non-ideal for quincy that could be related? | |
| 22:06:38 | gouthamr | i think the plugin ignores it, let me check | |
| 22:06:43 | dansmith | okay | |
| 22:06:59 | dansmith | https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/865315/11/.zuul.yaml#46 | |
| 22:08:43 | dansmith | okay just used for package repos anyway it looks like | |
| 22:08:51 | gouthamr | ack; we should get that opt out of the job because its confusing | |
| 22:08:53 | gouthamr | https://github.com/openstack/devstack-plugin-ceph/blob/563cb5deeb21815ce0c62fa30249e85e886c783a/devstack/lib/ceph#L981-L989 | |
| 22:09:43 | dansmith | ack, so I assume we're configuring a repo but just not installing anything from it on jammy, right? | |
| 22:10:02 | dansmith | I'll add it to my follow-on patch to add the optimization devstack vars | |
| 22:10:08 | dansmith | add .. the removal of it, I mean :) | |
| 22:10:13 | gouthamr | ack ty | |
| 22:10:20 | gouthamr | i may have messed this up in my last patch | |
| 22:10:21 | gouthamr | https://github.com/openstack/devstack-plugin-ceph/blob/563cb5deeb21815ce0c62fa30249e85e886c783a/devstack/lib/ceph#L1067-L1094 | |
| 22:10:47 | gouthamr | we're not invoking that method at all for ubuntu anymore.. | |
| 22:11:27 | dansmith | ack | |
| 22:11:45 | dansmith | well, there's a bunch of focal-specific stuff to clean up in there regardless | |
| 22:14:58 | dansmith | hmm, I think it's about to fail a test | |
| 22:15:11 | dansmith | been stopped for going on five minutes, assuming retrying a detach | |
| 22:15:38 | gouthamr | yes.. it was a head-scratcher; think vkmc and i noticed that our override to use download.ceph.com stopped working with focal at some point since ubuntu default-enabled the ubuntu ceph repos.. so dropping it made no difference, we ended up using/testing with the distro provided packages | |
| 22:15:43 | gouthamr | oh | |
| 22:16:19 | gouthamr | test_rebuild_server_with_volume_attached? | |
| 22:16:25 | dansmith | dunno yet | |
| 22:16:40 | gouthamr | ah | |
| 22:16:43 | dansmith | but it's in the rebuld group | |
| 22:16:46 | dansmith | yep | |
| 22:17:14 | dansmith | test_rebuild_server_with_volume_attached [430.305393s] ... FAILED | |
| 22:17:18 | dansmith | ugh | |
| 22:18:19 | dansmith | so the other variable here is the version of qemu and qemu's block-rbd driver are different in jammy of course, compared to what we've been testing | |
| 22:18:42 | dansmith | so could be a bug in one of those, especially since it's related to the detach in the guest | |
| 22:19:35 | gouthamr | ack; another thing to try would be to bump the ceph image to the latest quincy: https://github.com/openstack/devstack-plugin-ceph/blob/563cb5deeb21815ce0c62fa30249e85e886c783a/devstack/lib/cephadm#L32 | |
| 22:20:02 | dansmith | okay | |
| 22:20:17 | gouthamr | they've published v17.2.6 today, and v17.2.5 a month ago | |
| 22:21:01 | gouthamr | https://quay.io/repository/ceph/ceph?tab=tags | |
| 22:21:45 | dansmith | ack, I hate this sort of "version minesweeper" game.. if we're that sensitive to version, it feels like we're doing something wrong | |
| 22:22:39 | gouthamr | agreed; but since this stuff hasn't worked before on our ci, its worth a try | |
| 22:22:58 | dansmith | yeah for sure | |
| 22:23:18 | dansmith | volume resize passed | |
| 22:23:32 | dansmith | maybe we'll find that this is fewer fails or something | |
| 22:25:18 | dansmith | actually, that one didn't fail before | |
| 22:28:07 | gouthamr | okay, this may be good news? devstack-plugin-ceph-tempest-py3 is going to pass | |
| 22:28:34 | dansmith | no, really? | |
| 22:28:52 | dansmith | third recheck's a charm? | |
| 22:28:55 | gouthamr | :D | |
| 22:42:28 | dansmith | more fails on this cephadm job | |
| 22:42:48 | dansmith | so I guess it's not something fundamental, but maybe just massively less stable or we're hitting some race easier? | |
| 22:43:26 | dansmith | maybe it is memory-related and we're stressed more here | |
| 22:43:41 | dansmith | maybe I should try turning on the two optimizations here to see if that makes things more stable | |
| 22:43:54 | gouthamr | ++ | |