ianw | probably docker-compose has matured to the point that using system packages makes sense now | 00:40 |
---|---|---|
ianw | i imagine it was installed with pip originally because that was the only option | 00:41 |
clarkb | test results on https://review.opendev.org/c/opendev/system-config/+/882169 show a few things are unhappy with the older version unfortunately. We'll need to dig in and decide if we can live without those features | 00:44 |
*** amoralej|off is now known as amoralej | 07:34 | |
*** amoralej is now known as amoralej|lunch | 11:00 | |
*** amoralej|lunch is now known as amoralej | 12:09 | |
*** blarnath is now known as d34dh0r53 | 14:18 | |
clarkb | I'm going to hop into the gerrit community meeting (if I remembered it was happening earlier I could've added replication stuff to the agenda, oh well) and then I have some errands to run. But after that I'll dig into the docker-compose failures switching to the distro version | 14:58 |
fungi | thanks. i'm headed out shortly to a long lunch with friends but expect to be around again by 18:00 utc | 15:01 |
priteau | I typo'ed my meeting start command, if you want you can delete https://meetings.opendev.org/meetings/blkazar/ | 15:05 |
fungi | priteau: i'll try to remember to do it after lunch, have to authenticate to kerberos so i can delete that out of afs | 15:11 |
priteau | Thanks, no worries if you don't have time. | 15:13 |
clarkb | seems there is no one around to let us into the gerrit community meeting. Maybe I'll suggest we use meetpad for the next one or something | 15:14 |
fungi | let me guess... they're big on using google meet? | 15:17 |
opendevreview | Clark Boylan proposed opendev/system-config master: Align hound image build with Gitea's https://review.opendev.org/c/opendev/system-config/+/882328 | 15:17 |
fungi | okay, heading out, back in a few hours | 15:18 |
clarkb | o/ | 15:18 |
opendevreview | Clark Boylan proposed opendev/system-config master: Install docker-compose from the distro https://review.opendev.org/c/opendev/system-config/+/882169 | 15:36 |
clarkb | https://23.253.166.158:3081/opendev/system-config is the held gitea 1.19.3 node assuming we're ready to try that upgrade again after the fun yesterday | 15:40 |
clarkb | after a quick look it seems happy which isn't surprising since the delta to 1.19.2 is tiny | 15:40 |
opendevreview | Clark Boylan proposed opendev/system-config master: Install docker-compose from the distro https://review.opendev.org/c/opendev/system-config/+/882169 | 15:54 |
clarkb | I feel like I knew that was a bad variable name as I was typing it | 15:54 |
hashar | hi, I bet you already noticed but python urllib3 has released a new major version which breaks requests https://github.com/psf/requests/issues/6432 The fix is to add an upper bound constraint `urllib3<2` | 16:08 |
hashar | then I guess openstack already set up upper bound constraints on all dependencies so most probably none is affected :) | 16:08 |
Clark[m] | hashar: I spent all day dealing with that and associated fallout yesterday :( | 16:10 |
Clark[m] | I'm sure there will be more too unfortunately | 16:11 |
* hashar comforts Clark[m] with some cookies and donuts | 16:22 | |
hashar | those are no fun for sure | 16:22 |
*** amoralej is now known as amoralej|off | 16:55 | |
fungi | okay, back from lunch and catching up, i'll get to priteau's meeting log cleanup in a moment | 17:08 |
JayF | https://review.opendev.org/admin/repos/x/proliantutils,access does this indicate that literally nobody outside of the "All-Projects" access has access to this repo? | 17:10 |
JayF | Or am I unable to see it because I don't have access? | 17:10 |
JayF | I suspect it's that I can't see it because I don't have access... | 17:11 |
JayF | yeah, looks like it from a gander at project-config | 17:11 |
fungi | JayF: yes, oddly gerrit filters its acl display to only show access you have with the account you're logged in as | 17:13 |
fungi | they seem to consider showing you what permissions others have as a leak of sensitive information | 17:14 |
JayF | I have heard that argument made before :) | 17:14 |
JayF | either way, putting that knowledge in my pocket for the future :D | 17:14 |
fungi | so looking at the acls we put in project-config is a better bet | 17:14 |
fungi | okay, i apparently mis-remembered, meeting minutes and irc logs are just stored on a cinder volume attached to the eavesdrop server, not in afs. anyway, cleaned up the errant blkazar meeting directory now as priteau requested | 17:19 |
clarkb | the meeting schedule stuff is all on afs but the logs are locally mounted I think | 17:39 |
opendevreview | Clark Boylan proposed opendev/system-config master: Install docker-compose from the distro https://review.opendev.org/c/opendev/system-config/+/882169 | 17:41 |
clarkb | I'll figure out ansible one day | 17:41 |
opendevreview | Merged opendev/system-config master: Align hound image build with Gitea's https://review.opendev.org/c/opendev/system-config/+/882328 | 18:31 |
fungi | hound deploy just reported | 18:35 |
fungi | it'll be reloading for a while though, i expect | 18:35 |
fungi | currently returns "hound is not ready" when visited | 18:36 |
Clark[m] | It's about 5 minutes iirc | 18:36 |
Clark[m] | I want to say you can tail the log and see it do each repo | 18:36 |
Clark[m] | I think it is up now | 18:43 |
fungi | its's back, yep | 18:45 |
priteau | Thank you fungi | 19:26 |
fungi | since codesearch seems to still be working, i'll go ahead and reapprove the gitea upgrade | 19:57 |
fungi | the new held node still seems fine | 19:59 |
clarkb | fungi: thanks! I'm back and should be around for the rest of the day now and can keep an eye on it too | 20:18 |
fungi | cool, i meant to approve it a little sooner so it would be mostly ready once you were back | 20:19 |
clarkb | zuul has been very busy today | 20:20 |
clarkb | https://review.opendev.org/c/opendev/system-config/+/882169 passes CI now and gives us a very good idea of what is required for using distro docker-compose on top of getting everything to focal aready | 20:20 |
clarkb | in particular the pip module uninstalling distro installed docker-compose is something I'm not quite sure what to do about | 20:21 |
clarkb | I did what I think may work in the change but I'm open to better ideas too | 20:21 |
fungi | i would just leave the cruft, unless you want to force reinstall the distro package after pip uninstall | 20:54 |
clarkb | fungi: the problem with leaving it behind is then the pip version is used by default due to its location in $PATH and that I think would become extra confusing | 20:54 |
clarkb | so I'm really hopin we can remove t | 20:54 |
fungi | apt-get reinstall should fix it | 20:55 |
clarkb | I think what I've got should work everywhere but the gitea servers (where we seem to install both already | 20:55 |
clarkb | fungi: ya so the check I've got is probably sufficient | 20:55 |
clarkb | and if it doesn't work somewhere we can simply reinstall | 20:55 |
opendevreview | Merged opendev/system-config master: Update gitea to 1.19.3 https://review.opendev.org/c/opendev/system-config/+/877541 | 21:29 |
clarkb | here we go | 21:30 |
* fungi holds onto his seat | 21:30 | |
fungi | promote worked | 21:32 |
clarkb | gitea09 should be updated | 21:37 |
clarkb | and maybe a few others at this point | 21:37 |
clarkb | 09-12 are done. 13 is updating now | 21:38 |
fungi | Powered by Gitea Version: v1.19.3 | 21:40 |
fungi | seems to be working! | 21:41 |
clarkb | all 14 are done now | 21:41 |
clarkb | I think the last thing to check is if you can clone something and whether or not gerrit replication appears to work | 21:41 |
clarkb | cloning works for me | 21:43 |
fungi | yeah, working well here | 21:43 |
fungi | cloning nova is quite slow for me, but i think that's just the giant sucking sound that is zayo's atlanta network | 21:59 |
clarkb | I'm getting a similar bitrate to cloning system-config | 22:00 |
clarkb | ya I don't think nova is any worse than other repos for me | 22:01 |
opendevreview | Merged opendev/zone-opendev.org master: Remove old DNS servers https://review.opendev.org/c/opendev/zone-opendev.org/+/881935 | 22:05 |
fungi | Receiving objects: 100% (732881/732881), 156.16 MiB | 394.00 KiB/s, done. | 22:05 |
fungi | nearly 8 minutes start to finish | 22:05 |
ianw | ^^ I've removed the old servers too; so i'm pretty much calling the dns switch done | 22:05 |
fungi | mirror.dfw.rax got much better time, ~3 minutes to clone nova | 22:06 |
fungi | Receiving objects: 100% (614739/614739), 138.71 MiB | 1.16 MiB/s, done. | 22:06 |
clarkb | fungi: I got almost 3MBps | 22:08 |
clarkb | (but already closed the terminal window so can't copy the exact value anymore) | 22:08 |
clarkb | ianw: fwiw foundation reached out to the registrarand they gave us a very non confidence inspiring response so we asked them tochange nothing for glue records | 22:09 |
clarkb | so ya I think dns is done too. Thank you! | 22:09 |
ianw | ok, better something that works than not ... maybe 2024 will be the year of ipv6 :) | 22:09 |
fungi | mirror.iad.rax finished in about one minute | 22:10 |
fungi | Receiving objects: 100% (614739/614739), 138.64 MiB | 8.27 MiB/s, done. | 22:10 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!