| *** rlandy|ruck is now known as rlandy|ruck|out | 01:00 | |
| *** rlandy|ruck|out is now known as rlandy|out | 01:00 | |
| *** pojadhav|pto is now known as pojadhav | 05:02 | |
| *** akahat|PTO is now known as akahat | 06:02 | |
| *** akahat is now known as akahat|ruck | 06:30 | |
| *** ysandeep is now known as ysandeep|afk | 07:14 | |
| *** ykarel_ is now known as ykarel | 08:23 | |
| *** bhagyashris_ is now known as bhagyashris | 08:45 | |
| *** ysandeep|afk is now known as ysandeep | 09:16 | |
| opendevreview | chandan kumar proposed zuul/zuul-jobs master: Introduce iptables_package var https://review.opendev.org/c/zuul/zuul-jobs/+/822503 | 09:33 |
|---|---|---|
| opendevreview | chandan kumar proposed zuul/zuul-jobs master: Introduce iptables_package var https://review.opendev.org/c/zuul/zuul-jobs/+/822503 | 09:37 |
| *** ysandeep is now known as ysandeep|coffee | 10:15 | |
| *** rlandy|out is now known as rlandy|ruck | 11:02 | |
| *** dviroel_ is now known as dviroel|rover | 11:16 | |
| *** ysandeep|coffee is now known as ysandeep | 11:23 | |
| artom | Well then. It begins anew. | 13:29 |
| artom | ... but it begins poorly. Wrong channel :( | 13:29 |
| *** ykarel__ is now known as ykarel | 13:57 | |
| hrw | morning | 14:01 |
| hrw | is there an official way I can use to remove account 34360 and use marcin.juszkiewicz@linaro.org as email in 24072 account? | 14:01 |
| hrw | I created 34360 one during LP 2FA accident | 14:02 |
| hrw | I cannot just add other email and remove that one: "Error 409 (Conflict): Cannot remove e-mail 'marcin.juszkiewicz@linaro.org' which is directly associated with OPENID_SSO authentication" | 14:04 |
| opendevreview | Merged openstack/project-config master: Disable nodepool temporarily https://review.opendev.org/c/openstack/project-config/+/820369 | 14:10 |
| fungi | hrw: we've been working to clean up account conflicts and are down to a scant ~10% of the hundreds we started with. clarkb's been heading up that effort and can probably make the best recommendations for how (and when) to proceed with fixing that one | 14:16 |
| fungi | the good news is that the error you're getting is what prevents gerrit from making more conflicting accounts now that we've upgraded. the bad news is that its consistency checker won't let us clean up the existing conflicts easily (unless we can do them all in one batch), so we've been finding workarounds to be able to make incremental progress until we have the final batch as small as possible | 14:17 |
| opendevreview | Arnaud Morin proposed openstack/project-config master: Revert "Disable nodepool temporarily" https://review.opendev.org/c/openstack/project-config/+/823357 | 14:23 |
| hrw | fungi: thanks. | 14:30 |
| hrw | the only downside is that https://review.opendev.org/c/openstack/kolla/+/823398 is on other account than usual. but once it is in gerrit it is project's not mine (as 'everyone can push fixes') | 14:33 |
| *** ysandeep is now known as ysandeep|afk | 14:41 | |
| *** dviroel|rover is now known as dviroel|rover|lunch | 14:57 | |
| *** ysandeep|afk is now known as ysandeep | 15:21 | |
| *** dviroel|rover|lunch is now known as dviroel|rover | 15:48 | |
| clarkb | hrw: fungi: what we can do is "retire" the account (disable it and remove its preferred email setting) then the hack is to delete its openid entries via the rest api | 16:38 |
| clarkb | we can revert the retirement but not the deletion of openid entries easily | 16:38 |
| clarkb | I can walk someone through that later this week or do it myself next week likely | 16:38 |
| hrw | clarkb: 34360 account can go away and there is no rush | 16:59 |
| clarkb | hrw: ok, I'll get to it when I'm back to the normal day to day | 17:01 |
| hrw | clarkb: thanks! | 17:02 |
| opendevreview | Clark Boylan proposed opendev/system-config master: Update gitea to v1.15.9 https://review.opendev.org/c/opendev/system-config/+/823027 | 17:16 |
| clarkb | gitea made another bugfix release. That just shifts our upgrade change ahead to the latest version | 17:16 |
| clarkb | still no template deltas in git diff | 17:16 |
| jrosser_ | i've just started hacking on some centos-9 stream support in OSA, and am having some trouble with git-review | 17:18 |
| clarkb | are you using rsa keys? | 17:19 |
| jrosser_ | when trying to set up the git hooks it looks like the scp command is trying to use the sftp subsystem rather than scp, and it fails | 17:19 |
| jrosser_ | ah i've already fixed the keys thing | 17:19 |
| clarkb | so git review -s fails to scp the hook? | 17:20 |
| clarkb | do you have a traceback and related output? | 17:20 |
| jrosser_ | heres what happened out-of-the box https://paste.opendev.org/show/811914/ | 17:20 |
| clarkb | I wonder if this is part of the scp deprecation so centos is using sftp? | 17:21 |
| jrosser_ | i can't really see why it's choosing sftp, and passing -O to the scp command makes it work | 17:21 |
| clarkb | I'm on OpenSSH_8.8p1 and don't have this problem | 17:22 |
| clarkb | oh but with older openssl | 17:23 |
| clarkb | is scp actually sftp via alternatives or simmilar? | 17:24 |
| jrosser_ | they're different binaries | 17:25 |
| fungi | debug1: pledge: filesystem full | 17:25 |
| fungi | that's... odd | 17:25 |
| jrosser_ | i think thats badly worded "full access" | 17:27 |
| fungi | ahh | 17:27 |
| clarkb | I get that too. But I get `debug1: Sending command: scp -v -f hooks/commit-msg` | 17:27 |
| clarkb | so this apperas centos-9 specific to their build of openssh | 17:27 |
| clarkb | possibly coming to other distros but not there yet | 17:27 |
| fungi | i'm still on openssh 8.7p1, and the scp manpage indicates -O is on by default. is it off by default in 8.8p1? | 17:27 |
| jrosser_ | right from the word go (line 2 of the paste) it says sftp | 17:28 |
| clarkb | fungi: it may not log it as a default? | 17:29 |
| jrosser_ | my man page says " Use the legacy SCP protocol for file transfers instead of the SFTP protocol. Forcing the use of the SCP protocol may be necessary for servers that do not implement SFTP" for -O | 17:29 |
| fungi | yeah, mine says that too | 17:30 |
| fungi | and then says "This mode is the default." | 17:30 |
| fungi | maybe it's on by default on debian but not on other distros? | 17:30 |
| clarkb | mine says it is defalt too on tumbleweed | 17:30 |
| clarkb | but ya maybe centos changed that | 17:30 |
| jrosser_ | oh, in the HISTORY section, "Since OpenSSH 8.8 (8.7 in Red Hat/Fedora builds), scp has use the SFTP protocol for transfers by default." | 17:31 |
| jrosser_ | that'll be it | 17:31 |
| fungi | got it | 17:32 |
| clarkb | its annoying thatthey keep deciding to change defaults ahead of upstream | 17:32 |
| fungi | then yeah, for now we can stem the bleeding by adding -O in git-review and talking to gerrit/mina-sshd about sftp protocol support i guess | 17:32 |
| clarkb | fungi: I think we can fetch it via https instead/too but that might get weird if a .gitreview has ssh settings | 17:32 |
| clarkb | Another option would be to vendor the hook | 17:33 |
| fungi | might be nice to also get frickler's fix for parsing signed commits in git log merged while we're at it | 17:33 |
| fungi | but yeah, we've previously talked about vendoring a copy of the commit hook. i suppose we could make it try to fetch it anyway but treat an error there as not being a failure (just log a warning about it maybe) and then fall back on the vendored copy | 17:34 |
| clarkb | wow one-line doesn't mean one line amazing | 17:34 |
| clarkb | fungi: ++ | 17:35 |
| fungi | yes, that one's on the git devs, i'm afraid ;) | 17:35 |
| jrosser_ | i can make a patch for this, i think i've made it work | 17:39 |
| fungi | thanks! | 17:41 |
| opendevreview | Jonathan Rosser proposed opendev/git-review master: Force use of scp rather than sftp when installing commit hook https://review.opendev.org/c/opendev/git-review/+/823413 | 17:44 |
| fungi | tests should hopefully tell us if -O hasn't been around long enough | 17:45 |
| *** rlandy|ruck is now known as rlandy|ruck|lunch | 17:46 | |
| *** cloudnull1 is now known as cloudnull | 17:50 | |
| clarkb | is that why frickler's fix doesn't pass? | 17:58 |
| frickler | clarkb: yes, git on xenial is too old, will need to add some version check I guess | 17:59 |
| clarkb | we might be able to drop xenial at this point now that more things hvae updated. Would need to double check where we are still using it though | 18:01 |
| jrosser_ | looks like some version checking might be needed to use -O | 18:13 |
| jrosser_ | it's not there on my mac which is latest OSX | 18:13 |
| *** rlandy|ruck|lunch is now known as rlandy|ruck | 18:14 | |
| fungi | either version checking or try/except fallback | 18:16 |
| fungi | the matrix of possibilities will get even weirder once openssh drops scp -O (or the scp command for that matter), as well as once gerrit/mina-sshd add sftp protocol support | 18:17 |
| opendevreview | Jonathan Rosser proposed opendev/base-jobs master: Add nodeset for centos-9-stream https://review.opendev.org/c/opendev/base-jobs/+/823422 | 18:35 |
| fungi | clarkb: i can't recall, were we planning to meet this week? | 18:49 |
| Clark[m] | We weren't. It was last week that we said we'd play by ear but cancelled today | 18:51 |
| fungi | ahh, cool | 18:59 |
| fungi | i think openstack is hoping to get the skyline repos renamed into their namespace sometime soon, so revisiting renames at next week's meeting would likely be appreciated | 19:00 |
| jrosser_ | either of 821650 or 823422 adding a centos-9-stream nodeset would be cool if possible | 19:15 |
| fungi | ahh, thanks for the reminder! | 19:18 |
| jrosser_ | fungi: i think thats stacked on top of this as well https://review.opendev.org/c/opendev/base-jobs/+/821649 | 19:39 |
| fungi | oh, so it was. i thought i'd already approved that one weeks ago, but apparently not | 19:40 |
| opendevreview | Merged opendev/base-jobs master: Update Fedora latest nodeset to 35 https://review.opendev.org/c/opendev/base-jobs/+/821649 | 19:48 |
| opendevreview | Merged opendev/base-jobs master: Add 8-stream-arm64 and 9-stream nodesets https://review.opendev.org/c/opendev/base-jobs/+/821650 | 19:49 |
| fungi | jrosser_: ^ should be immediately available for use | 19:52 |
| *** dviroel|rover is now known as dviroel|rover|afk | 20:25 | |
| *** tosky_ is now known as tosky | 20:32 | |
| *** dviroel|rover|afk is now known as dviroel|rover | 21:46 | |
| *** dviroel|rover is now known as dviroel|out | 21:47 | |
| fungi | infra-root: it was brought up in a thread on openstack-discuss that the "tags" tab in gitea seems to only be populated with tags which were present in each repo at the time the database was created, but any subsequently pushed in via gerrit mirroring are not reflected. was this a known issue i'd somehow forgotten, or did gitea switch how they're generating the tags view at some point recently? | 21:58 |
| Clark[m] | It's been mentioned before but I don't think anyone has had time to debug it beyond "we really prefer you not rely on GitHub/gitea releases or tag tarballs | 23:47 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!