Friday, 2024-07-26

tonybclarkb, fungi: I expect it's a tipit issue but the cert on openstackid.org has expired.  `openssl s_client -servername openstackid.org -connect openstackid.org:443 </dev/null | grep -E '^[dv]'`01:59
tonybnotAfter=Jun 18 17:09:38 2024 GMT01:59
fungitonyb: thanks! i'll give them a heads up, but also id.openinfra.dev is the canonical name for the service these days and i think they only maintained the old hostname for backward-compatibility02:03
tonybThat makes sense.  I don't actually recall what click trail I followed that ended up at openstackid.org02:05
tonybwhatever it was, clearly needs to be updated02:05
fricklerclarkb: I found out that the hostId is not shown in the meta data, but in the nova API. "openstack server show gitea09.opendev.org -c hostId". there's a shared host for 09+14, another for 11+13, 10 and 12 are different05:33
fricklersee also https://docs.openstack.org/api-ref/compute/#id2105:52
fricklerthe nodepool builders are generating some CryptographyDeprecationWarning in their cron jobs for a couple days now. likely not critical yet, but maybe keep an eye on it?07:20
fricklerfwiw, abandoning a change in gerrit also had a noticable delay for me twice just now, though "only" about 5s instead of 1508:35
frickleralso, just for the record, the last nova CVE patch (for now) merged tonight, so nothing to look out for anymore. I've thus self-approved https://review.opendev.org/c/openstack/project-config/+/923509, waited long enough IMO08:38
opendevreviewMerged openstack/project-config master: Drop repos with config errors from x/ namespace  https://review.opendev.org/c/openstack/project-config/+/92350909:04
opendevreviewJan Gutter proposed zuul/zuul-jobs master: Update ensure-kubernetes with podman support  https://review.opendev.org/c/zuul/zuul-jobs/+/92497009:49
opendevreviewJan Gutter proposed zuul/zuul-jobs master: Update ensure-kubernetes with podman support  https://review.opendev.org/c/zuul/zuul-jobs/+/92497010:04
fricklergitea found out that I'm a Ghost, no more hiding https://opendev.org/openstack/tap-as-a-service/releases/tag/ocata-eol10:17
opendevreviewJan Gutter proposed zuul/zuul-jobs master: Update ensure-kubernetes with podman support  https://review.opendev.org/c/zuul/zuul-jobs/+/92497010:20
fricklerok, it is not only UI for gerrit, "git push -d gerrit stable/ocata" took about 20s showing "Remote: processing changes". I'll add "time" next time for more data, got a bunch to clean up still10:21
fricklerbut I think this is consistent with slow IO being the cause10:22
opendevreviewJan Gutter proposed zuul/zuul-jobs master: Update ensure-kubernetes with podman support  https://review.opendev.org/c/zuul/zuul-jobs/+/92497010:35
fricklerhmm, all further deletions for that repo (openstack/tap-as-a-service) went in < 3s10:37
opendevreviewJan Gutter proposed zuul/zuul-jobs master: Update ensure-kubernetes with podman support  https://review.opendev.org/c/zuul/zuul-jobs/+/92497011:22
frickleroh, nice, openstack-ansible-tests gate is running in an unnamed queue with window_size: infinity. now that's what I call headroom ;)11:46
fungifrickler: i think the cryptography deprecation message is also showing up in upstream zuul tests, there was some discussion of it in the zuul matrix too12:25
*** elodilles_ooo is now known as elodilles12:32
fungidiscussion there landed at https://github.com/paramiko/paramiko/pull/242112:42
fungiso i guess once there's a new paramiko release with that or a similar fix and we get newer nodepool container images with that version installed, it'll go away again12:44
fricklerah, thx for the pointer13:01
opendevreviewJan Gutter proposed zuul/zuul-jobs master: Update ensure-kubernetes with podman support  https://review.opendev.org/c/zuul/zuul-jobs/+/92497013:11
opendevreviewJan Gutter proposed zuul/zuul-jobs master: Update ensure-kubernetes with podman support  https://review.opendev.org/c/zuul/zuul-jobs/+/92497013:41
opendevreviewJan Gutter proposed zuul/zuul-jobs master: Update ensure-kubernetes with podman support  https://review.opendev.org/c/zuul/zuul-jobs/+/92497014:04
clarkbthe gitea ghost user appears to be the "we don't know who you are" stand in value15:34
clarkbinfra-root I just fowarded an email from the Works on Arm folks about the hosting for the current linaro arm cloud. TL;DR is that it is going away on August 10 (15 days notice...). I guess I'll respond to that email and let them know I can start widning down our usage of it next week15:51
fricklerah, I received some weird github notification the other day but forgot to mention it. also where did you forward the mail to?15:54
clarkbfrickler: I forwarded it to infra-root. And ya they posted initially to the original github issue saying to check our inboxes but I had no email so I emailed them directly. They didn't respond to that they just sent a new email but now we know15:55
clarkbin any case I'll write a response today thanking them for the use of the hardware and let them know I'll start sunsetting the usage on our side next week. Looks like the arm building is in osuosl so we don't need to redeploy anything shoudl be straightforward15:56
frickler+115:57
fungiany idea if osuosl's hardware is also on load from works-on-arm or was directly donated to them?16:03
fungii suppose Ramereth would know, if around16:04
fungis/on load/on loan/16:04
clarkbI suspect osuosls is compeltely separate. They have their own datacenters and everything16:04
clarkbits also different hardware iirc16:04
Ramereth[m]ours are separate16:04
fungiah, yeah maybe this is less of a concern for the hardware itself and more that arm/ampere doesn't want to keep paying equinix for the colo16:05
fungithanks for confirming Ramereth[m]!16:05
Ramereth[m]Do you need anything else from me?16:05
Ramereth[m]Are you needing more capacity?16:05
fungiRamereth[m]: we probably wouldn't turn down additional capacity, and will have less when the linaro environment disappears on us, but probably what we're really missing more is diversity of providers for arm resources so just adding more at osuosl wouldn't necessarily help with that16:06
*** dasm is now known as Guest132116:07
clarkbya I think we can see how we do after the shutdown and if things become painful we can talk about capacity. But for now I think we're good and in a wait and see mode16:07
Ramereth[m]fungi: okay good to know. I have regular contact with Ampere so I could bring that up if you'd like. Or if you have other questions/concerns related to their support16:07
fungiRamereth[m]: that's great to know, thanks! will definitely keep that in mind once we see what happens with this16:07
clarkbRamereth[m]: I may also end up at FOSSY or around FOSSY (not completely sure yet) if you wanted to discuss more in person16:08
fungiand also the existing resources there are very much appreciated, thanks again for all the hard work you and your crew put into it16:08
clarkbbut I don't think there is any urgency on your side. We continue to appreciate the resources you've made available to us!16:08
Ramereth[m]clarkb: great! let me know. It would be great to sync up16:10
opendevreviewClark Boylan proposed openstack/project-config master: Set linaro cloud's max servers to 0  https://review.opendev.org/c/openstack/project-config/+/92502916:38
clarkbI don't think we are in a rush to land ^ but figured I may as well push some of the changes16:39
opendevreviewCarlos Eduardo proposed openstack/project-config master: Implement manila-unmaintained-core group  https://review.opendev.org/c/openstack/project-config/+/92443018:05
JayFI'm having a really weird experience r/n with Gerrit. https://review.opendev.org/c/openstack/ironic-python-agent/+/924634 change, click on the -ipmi-direct-src job, click logs tab, navigate to controller/logs/ironic-bm-logs, click the first node-0 log. See it says "This logfile could not be found", but if you then go to the top right, and click "View log", you can navigate to the logs from the web directory of log files.18:45
fungithat's technically zuul not gerrit, but yeah, looking into it now18:49
fungiand the raw link when clicked tries to download rather than displaying in my browser, content type must be odd18:52
fungii wonder if it could be the url encoding of : throwing it off18:53
fungiJayF: do all log files exhibiting that behavior have : in their names, and only those as far as you've seen?18:56
JayFthis is literally the only time I've ever seen this behavior18:56
JayFbut I assumed the same you did18:56
fungithe zuul manifest for that build has raw : in the log filenames, the raw url has %-encoded : but i would assume the zuul dashboard code is smart enough to know that since it's what's providing both links18:59
Clark[m]The manifest is generated by the jobs not the dashboard I think19:08
Clark[m]So it could be a bug19:08
opendevreviewDan Smith proposed openstack/project-config master: Add openstack/os-test-images project under glance  https://review.opendev.org/c/openstack/project-config/+/92504319:09
opendevreviewDan Smith proposed openstack/project-config master: Add openstack/os-test-images project under glance  https://review.opendev.org/c/openstack/project-config/+/92504319:12
opendevreviewDan Smith proposed openstack/project-config master: Add openstack/os-test-images project under glance  https://review.opendev.org/c/openstack/project-config/+/92504321:05
opendevreviewClark Boylan proposed openstack/project-config master: Use the osuosl mirror for deb packages in image builds  https://review.opendev.org/c/openstack/project-config/+/92504822:00
opendevreviewClark Boylan proposed openstack/project-config master: Remove labels and diskimages from the linaro cloud  https://review.opendev.org/c/openstack/project-config/+/92504922:00
opendevreviewClark Boylan proposed openstack/project-config master: Remove linaro cloud from Nodepool  https://review.opendev.org/c/openstack/project-config/+/92505022:00
clarkbhttps://review.opendev.org/c/openstack/project-config/+/925048 and https://review.opendev.org/c/openstack/project-config/+/925029 should be safe to land at any time22:00
clarkbI can shephered the other changes through as these initial cleanups land22:01
clarkblooking at my calendar we did service coordinator nominations from Tuesday February 6 to Tuesday February 20 previously. 6 months from then is August 6 to August 20. I think we can do nominations from August 6 to August 20 then have an election from August 21 to August 28 if necessary. That also neatly ensures things are done before I'm traveling for the summit22:09
clarkbI'll put that on next week's meeting agenda and bring it up more formally there. But if there are concerns or suggestions sooner feel free to bring them up sooner (email or irc is fine)22:10
clarkbI've approved https://review.opendev.org/c/openstack/project-config/+/924430 for manila-unmaintained-core I don't think we need permission or anything for taht since openstack unmaintained branches are meant to be less controlled and open to whoever is willing to manage them22:40
clarkbthat said I'm not sure who should seed the new group. is if the change owner for that change cc fungi 22:40
opendevreviewMerged openstack/project-config master: Implement manila-unmaintained-core group  https://review.opendev.org/c/openstack/project-config/+/92443022:52
fungiclarkb: yeah, i'm not really sure either. i'd defer to the change proposer as well, but it's not entirely clear23:19

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!