| *** Guest3606 is now known as dasm | 01:48 | |
| *** dasm is now known as Guest4003 | 02:18 | |
| *** liuxie is now known as liushy | 02:30 | |
| opendevreview | Jake Yip proposed openstack/project-config master: Fork stackhpc/capi-helm-charts to openstack/magnum-capi-helm-charts https://review.opendev.org/c/openstack/project-config/+/918287 | 09:00 |
|---|---|---|
| dxld | hi! I'm starting a new project and I'm looking for (non-GH/GL) hosting. I really like opendev's idea of smushing gitea and gerrit together but it's not clear to me how to join as a new project? | 12:57 |
| dxld | so far I've been supporting sr.ht but I feel like it's not gone where I was hoping it would so my git forge budget is ready to be redirected too ;) | 12:59 |
| *** dhill is now known as Guest4184 | 13:30 | |
| frickler | dxld: did you see our documentation at https://docs.opendev.org/opendev/infra-manual/latest/gettingstarted.html and https://docs.opendev.org/opendev/infra-manual/latest/creators.html ? note that gitea is just "any git server" for us, the other main component besides gerrit is zuul doing CI/gating | 13:46 |
| dxld | I have but that seemed more geared towards openstack specifics? and I didn't spot a "here's how to sign up" thingy anyhow | 14:14 |
| dxld | and yeah using gitea just as a git frontend without all the other cruft is the bit I like :) | 14:15 |
| dxld | been looking for a CI that does this gating thing without knowing what to call it for years now too | 14:15 |
| dxld | so while I haven't looked at the details yet it sounds nice | 14:17 |
| *** Guest4003 is now known as dasm | 14:34 | |
| clarkb | dxld: https://docs.opendev.org/opendev/infra-manual/latest/creators.html#add-the-project-to-the-master-projects-list is the main bit for adding a project to our Gerrit. Lookingat the dosc there I wonder if we've got some of the order of operations not quite right (because we don't lead with how to create a project but instead have other stuff) | 15:08 |
| clarkb | dxld: that said, if you've never used gerrit before you amy want to push and review changes in our opendev/sandbox repo first to get a sense of whether or not it is something that you'll want to use | 15:09 |
| clarkb | if you decide that yes Gerrit's workflow works for you then ya creating the project as described in that link would be the very first step (then you can add all the other bits like zuul ci integration and irc bots etc) | 15:09 |
| dxld | I have used gerrit before in the coreboot project. just not very extensively | 15:34 |
| dxld | at this point I don't want to cause a buch of work on your side anyway. I'm still exploring and evaluating | 15:36 |
| clarkb | ya I think we do lack an explicit "this is how you create a project then tie into Zuul" document that can live a bit more on its own and that would be good to have | 15:38 |
| dxld | how does gitea source the list of projects/repos? is that all manually registered somehow? | 15:39 |
| clarkb | dxld: the CI/CD system runs a job when you update that project list that ensures all the repos are created in gitea. Then Gerrit replicates the repo content to gitea. | 15:41 |
| clarkb | Gitea is basically just a read only mirror for the git repos in Gerrit | 15:41 |
| dxld | right. that's what I figured ought to be happening :) | 15:42 |
| clarkb | on the held gerrit 3.8 test node I've set 2 of 3 changes to have topic "testtopic": I'll set the config for the topic limit to 1 then manually upgrade this node. This should check if the upgrade is impacted by the topic limit if any topics exceed the limit then we can also test what happens when you try to add another change to the topic | 16:21 |
| clarkb | https://158.69.71.178/q/topic:%22testtopic%22 | 16:21 |
| clarkb | https://review.opendev.org/c/opendev/system-config/+/912470 didn't promote the 3.9.4 image for some reason | 16:39 |
| clarkb | I thinkfor the same reason the etherpad image didn't promote :/ | 16:39 |
| clarkb | so anyway my test of gerrit upgrade is a bit behind on the gerrit version so I'll have to do this again I guess | 16:39 |
| clarkb | it does look like the change topic limit is fairly soft and is unlikely to create problems for us during my testing. I think we still retest with properly built images though and make sure that hasn't changed. I'll get a change up to address that shortly. Need to finish taking notes off the current held nodes | 16:47 |
| opendevreview | Clark Boylan proposed opendev/system-config master: Actually rebuild Gerrit images to get a new 3.9 image https://review.opendev.org/c/opendev/system-config/+/918333 | 16:53 |
| clarkb | tonyb: looks like all of the new-release topic changes are closed/merged/abandoned currently and based on my testing they shouldn't actually apply to the topic limit. Only open changes do | 16:55 |
| clarkb | and it doesn't look like we end up with 5k open at one time so I think we're good there | 16:56 |
| tonyb | okay that's good to know. | 16:57 |
| clarkb | (again this is with the wrong image/gerrit version but) the upgrade itself appears to go very quickly then we'll wait for online reindexing to finish before actually calling it done which I expect to take about half an hour in prod | 17:04 |
| clarkb | there are no db schema changes only index updates and from 3.8 to 3.9 they reindex online | 17:04 |
| clarkb | I won't bother testing downgrade stuff until images are sorted out because I won't feel confident in that testing if I do it with the older images anyway | 17:04 |
| clarkb | and my notes are in the etherpad from what I have learned. | 17:07 |
| tonyb | thanks clarkb | 17:48 |
| clarkb | fungi: I know yuo are not around and I think this is not urgent, but service-announce has 4 held messages. Two appear to be regular spam and the other two are generated by the receiving the spam? Makes me wonder if we have something misconfigured that creates that other two messages. I won't delete them unless the held queue gets huge so you can take a look later | 18:13 |
| clarkb | or maybe that is normal due to how the senders are set up | 18:14 |
| fungi | clarkb: senders on the spam were spoofed to be the same as the list address, seems like that's a pretty common tactic. we could consider banning the list addresses in the mta or something if it becomes an issue | 18:30 |
| clarkb | fungi: I see so just continue to drop both messages when clearing those queues? | 18:31 |
| fungi | yep | 18:31 |
| fungi | that's what i've been doing anyway | 18:32 |
| opendevreview | Merged opendev/system-config master: Allow overriding of meetpad PUBLIC_URL https://review.opendev.org/c/opendev/system-config/+/905510 | 19:34 |
| opendevreview | Merged opendev/system-config master: Use group var to define letsencrypt_certs https://review.opendev.org/c/opendev/system-config/+/905561 | 19:59 |
| opendevreview | Merged opendev/system-config master: Switch meetpad test servers to jammy https://review.opendev.org/c/opendev/system-config/+/905182 | 19:59 |
| tonyb | clarkb: I see that the backup volume on backup02 is > 90%. Any objection to me pruning it? | 20:15 |
| clarkb | tonyb: none from me | 20:16 |
| clarkb | and I see that email now | 20:17 |
| clarkb | going to start putting together a meeting agenda for tomorrow. Let me know if you want me to add anything or feel free to add content directly | 20:37 |
| tonyb | clarkb: Nothing from me. | 20:38 |
| tonyb | clarkb: backups pruned | 21:05 |
| clarkb | thanks! | 21:15 |
| clarkb | my meeting agneda edits are in now | 21:25 |
| tonyb | LGTM | 21:26 |
| opendevreview | Tony Breeds proposed opendev/zone-opendev.org master: Add DNS records for new Meetpad and JVB servers https://review.opendev.org/c/opendev/zone-opendev.org/+/918361 | 21:52 |
| opendevreview | Tony Breeds proposed opendev/system-config master: Add inventory records for new Meetpad and JVB servers https://review.opendev.org/c/opendev/system-config/+/918362 | 21:54 |
| clarkb | tonyb: before merging https://review.opendev.org/c/opendev/system-config/+/918362 we should consider putting the old servers in the emergency file | 22:09 |
| clarkb | tonyb: my main concern is that I'm not sure that multiple meetpad servers (whcih act as the main server for jitsi) can coordinate and operate in a cluster. So we don't want jvb02 and meetpad02 trying to sync with meetpad01 and jvb01? | 22:10 |
| clarkb | we want meetpad to make a new cluster independent of the old one then we'll update the meetpad.o.o dns record | 22:11 |
| * clarkb makes a note about this on the change | 22:11 | |
| tonyb | clarkb: Yup. Noted. | 22:17 |
| tonyb | I think the 02 servers will keep to themselves I'll login and check the 01 servers now have accurate xmpp_server settings if they do I htink they'll keep to themselves also | 22:18 |
| clarkb | tonyb: if we don't put them in the emergency file I think jvb01 and jvb02 may try to conncet to both meetpad01 and meetpad02 | 22:19 |
| clarkb | I don't know that the meetpads will try to connect to each other though as it is only the jvb's that are configured to look at group membership for that? | 22:19 |
| tonyb | clarkb: Ahh got it. yeah I'll double check but I agree adding the 01 servers to the emergency file is a good idea. | 22:21 |
| tonyb | I think I might be done for today. I just "did a dumb"(tm) and will need to resurrect my review chain from git .... perhaps my brain isn't working super well | 22:25 |
| clarkb | oh no. Not quite sure I undersatnd what went wrong but enjoy your evening | 22:26 |
| tonyb | nothing catastrophic. I just was working with a chain of changes locally and managed mess it up even though I was sure I was doing it right. | 22:46 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!