| *** mlavalle has quit IRC | 00:03 | |
| openstackgerrit | Merged opendev/system-config master: borg-backup-server: fix verification run https://review.opendev.org/c/opendev/system-config/+/781010 | 00:06 |
|---|---|---|
| openstackgerrit | Merged opendev/system-config master: kerberos-kdc: add database backups https://review.opendev.org/c/opendev/system-config/+/779891 | 00:07 |
| openstackgerrit | Merged opendev/system-config master: system-config-run-kerberos: run twice https://review.opendev.org/c/opendev/system-config/+/780979 | 00:07 |
| *** tosky has quit IRC | 00:14 | |
| ianw | opendev-prod-hourly has something seeminly stuck for 13 hr 23 min | 00:18 |
| ianw | it doesn't seem to have any jobs | 00:19 |
| fungi | so not blocking/blocked on a mutex | 00:19 |
| ianw | just a sort of zombie entry | 00:21 |
| openstackgerrit | Merged opendev/system-config master: Update nodepool launchers to focal in testing https://review.opendev.org/c/opendev/system-config/+/776981 | 00:42 |
| ianw | prometheanfire: https://29ad8d6a737f5d78d1ad-450f6f29734ae7f18d179491368c8872.ssl.cf2.rackcdn.com/781580/1/check/dib-nodepool-functional-openstack-gentoo-src/493470b/nodepool/builds/test-image-0000000003.log | 00:50 |
| ianw | * ERROR: sys-block/open-iscsi-2.0.875::gentoo failed (compile phase): | 00:51 |
| ianw | 2021-03-18 23:57:46.686 | * emake failed | 00:51 |
| prometheanfire | url not all on one line, but I'll take a look | 00:53 |
| prometheanfire | open-iscsi always seems to have problems and upstream tends to not care | 00:53 |
| ianw | https://zuul.opendev.org/t/openstack/build/493470b6ec89473f9506dd5e5451e3e8/logs is the job | 00:54 |
| prometheanfire | hmm, I'm guessing it's gcc-10 not liking open-iscsi | 00:57 |
| prometheanfire | without knowing | 00:57 |
| fungi | sounds like a safe bet | 00:58 |
| prometheanfire | 2021-03-18 23:57:46.686 | [01m[Kiscsiadm.c:859:56:[m[K [01;35m[Kwarning: [m[Karray subscript 2147483647 is outside the bounds of an interior zero-length array ‘[01m[Kstruct iscsi_stats_custom[0][m[K’ [[01;35m[K]8;;https://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html#index-Wzero-length-bounds-Wzero-length-bounds]8;;[m[K] | 00:59 |
| prometheanfire | heh, yep | 00:59 |
| prometheanfire | gonna build a image local to see if we can just use a newer version of open-iscsi | 01:01 |
| prometheanfire | hmm, built here | 01:15 |
| prometheanfire | 2021-03-19 01:15:25.533 | >>> Installing (3 of 3) sys-block/open-iscsi-2.0.875::gentoo | 01:16 |
| prometheanfire | oh, binary | 01:16 |
| prometheanfire | looks like I need to clean that up to match your fresh build :D | 01:16 |
| *** hamalq has quit IRC | 01:39 | |
| *** lbragstad has quit IRC | 02:06 | |
| prometheanfire | ok, error reproduced | 02:24 |
| prometheanfire | looks like the version we were using was remved from tree and it reverted to installing an earlier version | 02:27 |
| prometheanfire | so... it'll be a new release I think, I'll submit a review in a min | 02:28 |
| *** hemanth_n has joined #opendev | 02:57 | |
| *** hemanth_n has quit IRC | 03:37 | |
| *** hemanth_n has joined #opendev | 03:38 | |
| ianw | AnsibleUndefinedVariable: 'refstack_url' is undefined | 03:53 |
| ianw | from the refstack deploy job | 03:53 |
| openstackgerrit | Ian Wienand proposed opendev/system-config master: Rename refstack group variables https://review.opendev.org/c/opendev/system-config/+/781593 | 03:55 |
| *** JayF has quit IRC | 04:00 | |
| *** JayF has joined #opendev | 04:01 | |
| openstackgerrit | Matthew Thode proposed openstack/diskimage-builder master: update gentoo keywords to support gcc-10 https://review.opendev.org/c/openstack/diskimage-builder/+/781594 | 04:26 |
| *** ykarel has joined #opendev | 04:59 | |
| openstackgerrit | Ian Wienand proposed opendev/system-config master: Rename refstack group variables https://review.opendev.org/c/opendev/system-config/+/781593 | 05:02 |
| ianw | clarkb: if you could look at https://review.opendev.org/c/openstack/project-config/+/781002 ... that's currently applied manually so good to get it merged | 05:10 |
| openstackgerrit | Merged openstack/project-config master: Use nodejs8-publish-to-npm for monasca-grafana-datasource https://review.opendev.org/c/openstack/project-config/+/781536 | 05:17 |
| openstackgerrit | Merged openstack/project-config master: Normalize projects.yaml https://review.opendev.org/c/openstack/project-config/+/781019 | 05:21 |
| *** marios has joined #opendev | 06:02 | |
| *** ysandeep|away is now known as ysandeep | 06:15 | |
| *** elod is now known as elod_afk | 06:31 | |
| *** lpetrut has joined #opendev | 06:42 | |
| *** ykarel has quit IRC | 07:00 | |
| *** ykarel has joined #opendev | 07:02 | |
| *** sboyron has joined #opendev | 07:19 | |
| *** hashar has joined #opendev | 07:30 | |
| *** ysandeep is now known as ysandeep|lunch | 07:35 | |
| *** fressi has joined #opendev | 07:45 | |
| *** brinzhang0 has joined #opendev | 07:59 | |
| *** brinzhang_ has quit IRC | 08:01 | |
| *** rpittau|afk is now known as rpittau | 08:24 | |
| *** whoami-rajat_ has joined #opendev | 08:36 | |
| *** whoami-rajat_ is now known as whoami-rajat | 08:44 | |
| *** elod_afk is now known as elod | 08:48 | |
| *** andrewbonney has joined #opendev | 08:51 | |
| *** tosky has joined #opendev | 08:54 | |
| *** ykarel has quit IRC | 09:11 | |
| *** roman_g has joined #opendev | 09:23 | |
| *** ysandeep|lunch is now known as ysandee | 09:40 | |
| *** ysandee is now known as ysandeep | 09:56 | |
| *** ricolin has quit IRC | 10:01 | |
| *** hashar has quit IRC | 10:37 | |
| *** ykarel|lunch has joined #opendev | 10:45 | |
| openstackgerrit | Lee Yarwood proposed opendev/elastic-recheck master: Add query for bug #1920136 https://review.opendev.org/c/opendev/elastic-recheck/+/781621 | 10:48 |
| openstack | bug 1920136 in devstack "tempest-integrated-compute - c-sch/c-vol reporting insufficient free virtual space" [Undecided,In progress] https://launchpad.net/bugs/1920136 - Assigned to Lee Yarwood (lyarwood) | 10:48 |
| *** dtantsur|afk is now known as dtantsur | 10:52 | |
| *** ykarel|lunch is now known as ykarel | 11:06 | |
| dtantsur | ianw: hi! your patch does not work as it is, I think systemd-udev-settle is not triggered by default | 11:26 |
| dtantsur | I'm trying adding Wants on it | 11:26 |
| dtantsur | oh, hmm, and I think the path is wrong too | 11:34 |
| dtantsur | ianw: will you object if I update your patch after my testing? | 11:35 |
| openstackgerrit | Merged zuul/zuul-jobs master: Upgrade ansible-lint to 5.0 https://review.opendev.org/c/zuul/zuul-jobs/+/773245 | 11:37 |
| *** priteau has joined #opendev | 11:38 | |
| openstackgerrit | Dmitry Tantsur proposed opendev/glean master: Override NetworkManager to wait for udev-settle https://review.opendev.org/c/opendev/glean/+/781580 | 11:51 |
| openstackgerrit | Dmitry Tantsur proposed opendev/glean master: Allow disabling DHCP fallback https://review.opendev.org/c/opendev/glean/+/781500 | 11:52 |
| *** sboyron has quit IRC | 11:56 | |
| openstackgerrit | Dmitry Tantsur proposed openstack/diskimage-builder master: simple-init: allow disabling DHCP fallback https://review.opendev.org/c/openstack/diskimage-builder/+/781491 | 11:58 |
| *** hemanth_n has quit IRC | 11:59 | |
| *** roman_g has quit IRC | 12:03 | |
| *** roman_g has joined #opendev | 12:03 | |
| *** roman_g has quit IRC | 12:03 | |
| *** roman_g has joined #opendev | 12:04 | |
| *** roman_g has quit IRC | 12:04 | |
| *** roman_g has joined #opendev | 12:05 | |
| *** roman_g has quit IRC | 12:05 | |
| dtantsur | ianw, clarkb, I hope these three patches ^^^ solve my struggles finally | 12:06 |
| *** ykarel has quit IRC | 12:06 | |
| *** ykarel has joined #opendev | 12:07 | |
| *** roman_g has joined #opendev | 12:07 | |
| *** roman_g has joined #opendev | 12:08 | |
| *** roman_g has quit IRC | 12:08 | |
| *** lbragstad has joined #opendev | 12:22 | |
| *** ysandeep is now known as ysandeep|afk | 12:27 | |
| *** zbr is now known as zbr|rover | 12:30 | |
| *** ykarel has quit IRC | 12:35 | |
| *** ykarel has joined #opendev | 12:35 | |
| openstackgerrit | Dmitry Tantsur proposed opendev/glean master: Add Python 3.6 to supported versions https://review.opendev.org/c/opendev/glean/+/781674 | 12:57 |
| *** roman_g has joined #opendev | 13:00 | |
| *** lbragstad_ has joined #opendev | 13:25 | |
| *** fressi_ has joined #opendev | 13:26 | |
| *** ysandeep|afk is now known as ysandeep | 13:27 | |
| *** dhellmann_ has joined #opendev | 13:28 | |
| *** gothicserpent_ has joined #opendev | 13:30 | |
| *** brinzhang0 has quit IRC | 13:30 | |
| *** calcmandan_ has joined #opendev | 13:30 | |
| *** brinzhang0 has joined #opendev | 13:31 | |
| *** lbragstad has quit IRC | 13:33 | |
| *** fressi has quit IRC | 13:33 | |
| *** irclogbot_3 has quit IRC | 13:33 | |
| *** dhellmann has quit IRC | 13:33 | |
| *** gothicserpent has quit IRC | 13:33 | |
| *** openstackgerrit has quit IRC | 13:33 | |
| *** calcmandan has quit IRC | 13:34 | |
| *** fressi_ is now known as fressi | 13:34 | |
| *** dhellmann_ is now known as dhellmann | 13:34 | |
| *** irclogbot_1 has joined #opendev | 13:35 | |
| *** mlavalle has joined #opendev | 13:58 | |
| *** hashar has joined #opendev | 14:06 | |
| *** sboyron has joined #opendev | 14:18 | |
| dtantsur | Folks, I see a very annoying (and likely new) behavior in Gerrit UI: when you click "Reply", the focus on the "Reviewers" field, not on the message | 14:26 |
| fungi | that's interesting | 14:46 |
| fungi | you haven't noticed that before today? | 14:46 |
| dtantsur | No. I wasn't paying attention, but given how annoying I find it now, I would have noticed. | 14:47 |
| fungi | i'm wondering if a behavior change crept into their stable-3.2 branch. we continuously build and deploy images of that | 14:49 |
| *** gothicserpent_ is now known as gothicserpent | 14:51 | |
| lourot | o/ I'm getting an openstack-zuul-jobs-linters failure that isn't caused by my review I think, what do you think? https://review.opendev.org/c/openstack/project-config/+/781735 | 14:53 |
| *** ykarel has quit IRC | 14:53 | |
| iurygregory | lourot, yeah I just noticed the same problem on my patch https://review.opendev.org/c/openstack/project-config/+/775244 =) | 14:57 |
| *** roman_g has quit IRC | 15:02 | |
| *** roman_g has joined #opendev | 15:03 | |
| *** roman_g has quit IRC | 15:03 | |
| clarkb | fungi: dtantsur: note it isn't truly continuous, we have to land changes to our gerrit image files which causes rebuilds. I think it has been a few weeks since our last rebuild | 15:03 |
| fungi | okay, so if it's a new behavior in gerrit, it's been this way at least a few weeks | 15:03 |
| dtantsur | hmm, maybe I've just noticed.. I mostly use gertty | 15:04 |
| fungi | yeah, i use gertty almost exclusively | 15:04 |
| clarkb | lourot: iurygregory: ../../zuul/zuul-jobs/roles/revoke-sudo/tasks/main.yaml:10: [E302] rm used in place of argument state=absent to file module <- I wonder why that job is linting the zuul-jobs content | 15:05 |
| clarkb | I suspect something changed in zuul-jobs causing those failures, but ozj shouldn't care about zuul-jobs content | 15:05 |
| fungi | https://review.opendev.org/c/zuul/zuul-jobs/+/773245 Upgrade ansible-lint to 5.0 | 15:05 |
| fungi | that just merged | 15:05 |
| clarkb | fungi: this is running ansible-lint on ozj though and examining zuul-jobs files | 15:06 |
| clarkb | er project-config I Guess | 15:06 |
| clarkb | which shouldn't try and lint zuul-jobs | 15:07 |
| fungi | yeah, just wondering if ansible-lint 5.0 changed how files are included/excluded | 15:08 |
| *** ysandeep is now known as ysandeep|dinner | 15:08 | |
| clarkb | but if the job ran on project-config it shouldn't matter what version of ansile-lint zuul-jobs uses | 15:08 |
| *** ykarel has joined #opendev | 15:08 | |
| *** rpittau is now known as rpittau|afk | 15:08 | |
| clarkb | I think I see why it looks at zuul-jobs though. we set ANSIBLE_ROLES_PATH={env:ANSIBLE_ROLES_PATH:../zuul-jobs/roles:../openstack-zuul-jobs/roles:roles} | 15:09 |
| clarkb | I feel like this is a bug in ansible-lint | 15:09 |
| clarkb | this would effectively force all of your third party roles to lint clean | 15:10 |
| fungi | i'm wondering if it changed a default ansible-lint version in the job we're running | 15:10 |
| clarkb | which just isn't a sustainable approach | 15:10 |
| clarkb | project-config seems to set ansible-lint>=4.2.0,<5 | 15:10 |
| fungi | oh, yeah. and now i see what you're saying... why are we linting ../../zuul/zuul-jobs/roles/ in a project-config change? | 15:11 |
| clarkb | exactly | 15:11 |
| zbr|rover | i can give some insights on v5 | 15:11 |
| clarkb | v5 is not relevant | 15:11 |
| clarkb | we are using 4.3.7 | 15:11 |
| *** lbragstad_ is now known as lbragstad | 15:11 | |
| clarkb | the problem is we set ANSIBLE_ROLES_PATH and ansible-lint lints the files in the roles path in addition to the specific files we have told it to lint. I don't think this is a viable approach ebcause you cannot expect third party roles to adhere to the same linter rules as your ansible | 15:12 |
| clarkb | (it is possible that ansible-lint is finding those roles through some other mechanism but the one that seems most likely is the ANSIBLE_ROLES_PATH var we set) | 15:13 |
| clarkb | if v5 fixes this behavior we could upgrade to it, but I don't think v5 is involved in causing the behavior we see | 15:14 |
| zbr|rover | likely it does, but you can also add `../../zuul` to ignore path, it should work with older versions | 15:15 |
| zbr|rover | on v5, if you do not feed it any arguments, it will only lint what git ls-files reports, ignoring untracked stuff, one benefit being that the chance to need excludes is considerably lower. | 15:17 |
| clarkb | we lost openstackgerrit but remote: https://review.opendev.org/c/openstack/project-config/+/781774 Stop linting zuul-jobs from here | 15:21 |
| clarkb | I've restarted gerrit bot | 15:24 |
| *** chkumar|ruck is now known as raukadah | 15:24 | |
| lourot | thanks for the quick reaction! | 15:25 |
| *** fressi has quit IRC | 15:26 | |
| fungi | clarkb: so what's the theory on the regression? we've been linting the content of the zuul-jobs repo in our project config changes all along, but some of the content in zuul-jobs changed and the version of ansible-lint we're running no longer likes it? | 15:29 |
| clarkb | fungi: yes, I think zuul-jobs was following the project-config rules until recently and now linting on project-config complains | 15:31 |
| clarkb | I suspect something changed either in the linter or in zuul-jobs | 15:31 |
| fungi | probably the changes made to zuul-jobs to satisfy ansible-lint 5 in that case, given the timing | 15:32 |
| *** openstackgerrit has joined #opendev | 15:32 | |
| openstackgerrit | Clark Boylan proposed opendev/system-config master: Upgrade gitea to 1.13.4 https://review.opendev.org/c/opendev/system-config/+/780923 | 15:32 |
| openstackgerrit | Clark Boylan proposed opendev/system-config master: Force gitea to fail for human verification https://review.opendev.org/c/opendev/system-config/+/781776 | 15:32 |
| zbr|rover | something is weird as zuul-jobs errors should have never ended up being reported and based on the existing config which already has ../../opendev but not zuul jobs. | 15:34 |
| clarkb | zbr|rover: zuul jobs are in ../../zuul/zuul-jobs not ../../opendev | 15:34 |
| clarkb | I suspect we ran into this with opendev/base-jobs as well | 15:35 |
| clarkb | fixed it there and moved on. And now we are seeing it for zuul-jobs too | 15:35 |
| fungi | yeah, that seems increasingly likely then | 15:35 |
| clarkb | I've set a hold on the gitea job for 781776 | 15:36 |
| clarkb | we can use that to check on gitea upgrade results | 15:36 |
| clarkb | (though it does a fresh install not an upgrade) | 15:36 |
| zbr|rover | if any of you want me to make CR to upgrade the linter on another repo, let me know. zuul-jobs was an interesting experience, but I endup migrating both custom rules to the linter, making much easier to maintain. | 15:39 |
| clarkb | if we think that v5 will fix this issue then upgrading is probably worth looking at | 15:42 |
| *** sboyron has quit IRC | 15:42 | |
| clarkb | fungi: is there a gerritbot change to ignore messages for commits that are too long? | 15:43 |
| clarkb | (also are people writing commit messages that are too long frequently because we aren't helping them learn to write better ones? I wonder if we shoudl also try to reach out to people doing that0 | 15:43 |
| fungi | clarkb: no, not yet, i wanted to take it as an opportunity to try to fix the reconnect behavior as well | 15:43 |
| toomer | clarkb: Do you know if it's posible to migrate Gerrit user from one externalId(OAuth) to another externalID(SAML) | 16:04 |
| clarkb | toomer: yes, you create an externalid in All-Users:refs/external-ids with the right hashed path for the externalid value and ensure the accountId is set properly in that file. You need to reindex accounts if you do that change with gerrit offline | 16:05 |
| clarkb | toomer: the docs are not good on this stuff and I've been trying to think of ways to improve them as I work through our gerrit account consistency cleanups (as its a similar process) | 16:06 |
| clarkb | if you are using pre notedb gerrit for accounts then you just update the account_external_ids db table | 16:06 |
| toomer | I'm using NoteDB and woneering if something like that is even posible | 16:07 |
| toomer | There is not much documentation about this proccess on Internet | 16:07 |
| toomer | Do you have any notes that could help me here ? | 16:07 |
| clarkb | toomer: ya there was a thread that andrew grimburg wrote on for account cleanups but the processes are similar. Let me find that for you | 16:08 |
| toomer | Is that mean I will need to create an externalID for existing user with right hash but without password | 16:10 |
| openstackgerrit | Merged openstack/project-config master: Stop linting zuul-jobs from here https://review.opendev.org/c/openstack/project-config/+/781774 | 16:11 |
| clarkb | toomer: yes, each openid/oath/saml/email/etc is a separate exteranl id | 16:11 |
| clarkb | that externalid is located in a file named after a hashed value for the id itself | 16:11 |
| clarkb | then in that file it has data like which account does this id belong to | 16:11 |
| toomer | Even when the same emial address is used | 16:11 |
| toomer | Just the authentication protocol change: OATH --> SML | 16:11 |
| toomer | SAML | 16:12 |
| clarkb | toomer: https://groups.google.com/g/repo-discuss/c/4mbykY3oS1o/m/bcDGeJq5AQAJ that is my original thread, https://groups.google.com/g/repo-discuss/c/tZ1tYQwbeLY/m/xSZhIQ20EQAJ is Andrew's older info on how to address it | 16:12 |
| clarkb | toomer: yes, since the files are keyed by the exteranlid value not their attributes (like email or account id) | 16:12 |
| clarkb | toomer: also note that you can't have different accounts with the same emails in external ids. THis is the problem I am not cleaning up | 16:12 |
| toomer | clarkb: Thanks a LOT again ! | 16:12 |
| clarkb | I have ~545 conflicts to address on that problem :( | 16:12 |
| clarkb | but that is down from like 750 conflicts in total so slow progress | 16:13 |
| *** ysandeep|dinner is now known as ysandeep | 16:13 | |
| toomer | I will take a look on those note and see if I can make that worked | 16:13 |
| fungi | toomer: yeah, to reiterate, the main thing to keep in mind is that you need to know the new id you're migrating to ahead of time so you can pre-add it to the account | 16:14 |
| *** brinzhang0 has quit IRC | 16:14 | |
| toomer | What you mean by new ID ? | 16:14 |
| clarkb | dtantsur: fungi: I just rechecked the project-config changes above since their fix has landed. When I did so I copy pasted the recheck comment and it went straight into the message not the reviewers field. The issue is probably more subtle than the focus is always wrong | 16:14 |
| fungi | basically when the user logs in with the new id, you want gerrit to find there's already a record for that and use it, not rely on gerrit to create the id | 16:14 |
| *** brinzhang0 has joined #opendev | 16:15 | |
| clarkb | in your case sounds like the new ID would be the new SAML ID | 16:15 |
| toomer | fungi Exactly, this is what I need | 16:15 |
| dtantsur | yeeah.. I have 4 pages with Gerrit open, I see the issue on 2 of them | 16:15 |
| openstackgerrit | Aurelien Lourot proposed openstack/project-config master: Mirror new Magnum and Manila charms to GitHub https://review.opendev.org/c/openstack/project-config/+/781735 | 16:15 |
| fungi | right, that. basically know the saml ids of your users in advance, and then add those to gerrit manually associated with their accounts. gerrit will almost certainly not do the right thing if you rely on it to just notice the e-mail addresses are the same or something | 16:16 |
| clarkb | lourot: note you didn't need to rebase. Zuul will do that for you | 16:16 |
| dtantsur | only on one after refresh | 16:16 |
| clarkb | dtantsur: those are the best bugs | 16:16 |
| *** ykarel has quit IRC | 16:16 | |
| dtantsur | it is with https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/781745 can it be because it has no reviewers yet? | 16:16 |
| *** ykarel has joined #opendev | 16:17 | |
| toomer | Looks like you are Gerrit experts on this stage. I assume you have lots of diffrent kind of Gerrit issues in OpenDev | 16:17 |
| clarkb | dtantsur: if I reivew that one I get a cursor in the message box not reviewers list. I don't think that is it, but maybe it is a component of the issue? | 16:17 |
| dtantsur | clarkb: maybe it has to be your review? | 16:17 |
| clarkb | dtantsur: ya maybe that is it let me see if I have one I can check like that | 16:18 |
| clarkb | toomer: I don't know that we are experts, we certainly try though :) the upstream mailing list is pretty responsive though a lot of issues around account management have been reported over time and not a ton of improvements have been made that I can see | 16:18 |
| clarkb | dtantsur: yup if I review https://review.opendev.org/c/opendev/system-config/+/781776 my cursor goes to the reviewers | 16:19 |
| clarkb | now let me add fungi and see if it fixes it (sorry fungi if that generates eamil noise for you) | 16:19 |
| toomer | clarkb: Yes, the user managment in Gerrit is none exsitent | 16:19 |
| clarkb | dtantsur: fungi yup now my cursor goes to the message field. NEAT | 16:20 |
| dtantsur | heh, it may even make sense in some workflows | 16:20 |
| fungi | i have lots of e-mail noise already. you're spitting into the ocean there | 16:20 |
| dtantsur | next time I tag you on a review, I can claim that gerrit forced me :) | 16:20 |
| *** ykarel is now known as ykarel|away | 16:29 | |
| *** hashar has quit IRC | 16:33 | |
| *** lpetrut has quit IRC | 16:47 | |
| *** ysandeep is now known as ysandeep|away | 16:50 | |
| openstackgerrit | Jeremy Stanley proposed openstack/project-config master: Use Node 10 to release monasca-grafana-datasource https://review.opendev.org/c/openstack/project-config/+/781826 | 16:55 |
| openstackgerrit | Clark Boylan proposed opendev/system-config master: Force gitea to fail for human verification https://review.opendev.org/c/opendev/system-config/+/781776 | 16:58 |
| *** marios is now known as marios|out | 17:17 | |
| *** dtantsur is now known as dtantsur|afk | 17:21 | |
| *** ykarel|away has quit IRC | 17:31 | |
| *** marios|out has quit IRC | 17:32 | |
| openstackgerrit | Clark Boylan proposed opendev/system-config master: Set up gitea image provides and requires for gating https://review.opendev.org/c/opendev/system-config/+/781830 | 17:35 |
| clarkb | corvus: fungi ^ can you double check the assertions in that change? but i think that may address some weirdness with testing of the new gitea version that I have noticed. | 17:35 |
| clarkb | I can also redo the stack for new gitea to exercise ^ if we land it | 17:35 |
| corvus | clarkb: any job ordering changes needed for that? | 17:37 |
| corvus | (i'm checking too right now; but just wondering if you'd gone through those already) | 17:38 |
| corvus | looks like run-gitea depends on registry, build-image-gitea and build-image-haproxy-statsd; so i think that's all good | 17:39 |
| clarkb | ya I think that was already in place so it worked fine at a buildset level | 17:39 |
| clarkb | but not across changes | 17:39 |
| fungi | yeah, i think the prior behavior would have been to use the image from that same buildset if there was one, otherwise use the published image? | 17:41 |
| clarkb | yes I think that is what I was observing | 17:41 |
| *** hamalq has joined #opendev | 17:41 | |
| fungi | so this should let you consume an image built in a depended-on change | 17:41 |
| clarkb | or even a git parent child change | 17:41 |
| fungi | er, yeah | 17:41 |
| fungi | explicit or implicit change dependency, either one | 17:42 |
| fungi | also change ahead of it in a dependency queue | 17:42 |
| fungi | any of the above | 17:42 |
| *** sboyron has joined #opendev | 17:43 | |
| *** andrewbonney has quit IRC | 17:50 | |
| clarkb | https://198.72.124.60:3081/opendev/system-config that was with the forced rebuild in the child job | 18:28 |
| clarkb | looks fine to me. I would like to redo it though with 781830 in place | 18:29 |
| *** hamalq has quit IRC | 18:36 | |
| *** hamalq has joined #opendev | 18:36 | |
| openstackgerrit | Merged openstack/project-config master: Use Node 10 to release monasca-grafana-datasource https://review.opendev.org/c/openstack/project-config/+/781826 | 18:41 |
| fungi | clarkb: so i *think* i see the problem with https://opendev.org/opendev/gerritbot/src/branch/master/gerritbot/bot.py#L135 | 18:59 |
| fungi | we should be calling https://github.com/jaraco/irc/blob/acd2451666b26ee9e183f673fd9b3689ff401f91/irc/client.py#L212 | 18:59 |
| *** whoami-rajat has quit IRC | 19:00 | |
| fungi | the factory in connection.py doesn't have a reconnect() method | 19:01 |
| fungi | we're subclassing https://github.com/jaraco/irc/blob/acd2451666b26ee9e183f673fd9b3689ff401f91/irc/client.py#L1119 which doesn't expose that | 19:07 |
| fungi | anyway, about to start driving again, so i'll pick this up again when i'm back at the house | 19:08 |
| fungi | ahh, okay the method is actually defined at irc.client.ServerConnection.reconnect() and the irc.client.SimpleIRCClient we subclass has access to the irc.client.Reactor methods where SimpleIRCClient.connection is an instance of ServerConnection | 19:47 |
| fungi | so calling self.reconnect() from gerritbot.bot.GerritBot is legitimate, that method should exist, it just isn't properly reconnecting for some reason. i wonder if the sasl/ssl mix-in classes are involved | 19:50 |
| fungi | like maybe the saved connection context isn't actually including the auth or encryption | 19:50 |
| fungi | oh, OR it could be that we're just failing to clear the joined_channels structure when we reconnect and so it assumes it's already in channels and doesn't bother rejoining them | 19:51 |
| fungi | i wonder if simply inserting self.joined_channels = {} before the self.connection.reconnect() would fix this | 19:52 |
| fungi | this may turn out to be a regression from when we added the dynamic joining and unjoining | 19:54 |
| openstackgerrit | Gage Hugo proposed opendev/irc-meetings master: Update security sig frequency and chair https://review.opendev.org/c/opendev/irc-meetings/+/781902 | 20:09 |
| *** sboyron has quit IRC | 20:18 | |
| openstackgerrit | Merged opendev/irc-meetings master: Update security sig frequency and chair https://review.opendev.org/c/opendev/irc-meetings/+/781902 | 20:40 |
| clarkb | fungi: oh neat | 21:05 |
| clarkb | fungi: also doing an explicit disconnect, cleanup, then connect seems reasonable if that reads better | 21:05 |
| fungi | i think that's all reconnect does anyway, the connect method is supposed to replace the current connection, and i'm starting to think that's working fine. freenode logs that the client quit, the client doesn't log any errors | 21:10 |
| clarkb | does anyone know how https://review.opendev.org/c/opendev/system-config/+/780989/1/testinfra/test_nodepool.py and https://review.opendev.org/c/opendev/system-config/+/776981/2/testinfra/test_nodepool.py avoided merge conflicting? | 21:16 |
| clarkb | does (j)git see they both produce the same end state so doesn't consider it a conflict? | 21:16 |
| clarkb | I don't think this is wrong in this case, mostly just surprising and wanted to see if anyone knew how that happens in a (j)git context | 21:17 |
| fungi | i think so. at least that's how (c)git works | 21:17 |
| fungi | if i cherry-pick a commit which makes some changes which are already present in my tree, those changes are skipped silently | 21:18 |
| fungi | i rarely call git merge myself, but expect it uses the same conflict resolution mechanisms | 21:18 |
| fungi | same conflict resolution mechanisms as cherry-pick i mean | 21:19 |
| clarkb | gotcha | 21:19 |
| clarkb | I put those test update changes up separate to the actual flip over cahnges because the test only chagnes were useful to ensure there weren't any problems upfront | 21:20 |
| clarkb | but then it was awkward to try and do the coordination of the changes in that context | 21:20 |
| clarkb | anyway shouldn't be a problem just thought it surprising | 21:20 |
| *** openstackgerrit has quit IRC | 21:22 | |
| fungi | i'm fiddling with gerritbot btw, in case anyone notices exit messages. i'll keep it running | 21:24 |
| *** openstackgerrit has joined #opendev | 21:35 | |
| openstackgerrit | Jeremy Stanley proposed opendev/gerritbot master: DNM: Triggering a gerritbot event https://review.opendev.org/c/opendev/gerritbot/+/781915 | 21:35 |
| *** openstackgerrit has quit IRC | 21:35 | |
| *** openstackgerrit has joined #opendev | 21:37 | |
| *** openstackgerrit has joined #opendev | 21:37 | |
| openstackgerrit | Jeremy Stanley proposed opendev/gerritbot master: DNM: Triggering a gerritbot event https://review.opendev.org/c/opendev/gerritbot/+/781915 | 21:37 |
| fungi | clarkb: yep! that was precisely it. fix for that on the way, along with one to ignore that particular exception and not reconnect when it's seen | 21:38 |
| clarkb | cool | 21:39 |
| openstackgerrit | Jeremy Stanley proposed opendev/gerritbot master: Explicitly clear joined channels when reconnecting https://review.opendev.org/c/opendev/gerritbot/+/781919 | 21:51 |
| openstackgerrit | Jeremy Stanley proposed opendev/gerritbot master: Don't reconnect on MessageTooLong exceptions https://review.opendev.org/c/opendev/gerritbot/+/781920 | 21:51 |
| fungi | gotta love tracking down bugs which result in nothing being logged as amiss, and discovering that's because nothing actually was as far as the script was concerned | 21:52 |
| fungi | infra-root: ^ i think a lot of our mysterious "dunno netsplit maybe?" events where gerritbot never came back was that, it's been happening upwards of a few times a week | 21:53 |
| fungi | hopefully that's time well-spent since it should save us time constantly restarting gerritbot | 21:53 |
| clarkb | ++ any idea if there are users we need to help with commti writing suggestions (I would guess your log inspection would expose that maybe?) | 21:54 |
| fungi | not really... it seems to be a mix of new users who don't realize commit messages need a separate subject, and users who just forget to add a blank line after the subject | 21:55 |
| fungi | on projects where gerrit is enforcing a max commit message subject length this probably never comes up | 21:55 |
| corvus | wow | 21:56 |
| fungi | the container on eavesdrop is currently running with 781919 hand patched in | 21:57 |
| fungi | and i exercised it with a couple of do-not-merge changes just to make sure it behaved as suspected | 21:57 |
| openstackgerrit | Jeremy Stanley proposed opendev/gerritbot master: Don't reconnect on MessageTooLong exceptions https://review.opendev.org/c/opendev/gerritbot/+/781920 | 22:05 |
| fungi | added a log message | 22:05 |
| clarkb | fungi: do you know why testing failed on ps1? | 22:07 |
| fungi | i do not, saw the py37 job failed but haven't checked the log yet | 22:08 |
| fungi | curious to see if it also fails on the other change | 22:08 |
| fungi | ahh, nope, it passed on the first change | 22:08 |
| *** dpawlik has quit IRC | 22:11 | |
| *** gothicserpent has quit IRC | 22:11 | |
| *** lbragstad has quit IRC | 22:11 | |
| *** akahat has quit IRC | 22:11 | |
| fungi | aha, my mistake ;) | 22:12 |
| fungi | ModuleNotFoundError: No module named 'irc.client.MessageTooLong'; 'irc.client' is not a package | 22:12 |
| clarkb | oh you need from irc.client import MessageTooLong | 22:12 |
| fungi | jftr, that's the change i didn't test ;) | 22:12 |
| clarkb | or just import irc.client | 22:12 |
| fungi | yeah | 22:12 |
| *** brinzhang_ has joined #opendev | 22:12 | |
| fungi | i need to see if it has side effects first | 22:12 |
| *** lbragstad_ has joined #opendev | 22:12 | |
| *** Dmitrii-Sh8 has joined #opendev | 22:12 | |
| *** dpawlik9 has joined #opendev | 22:12 | |
| fungi | looks like it's all function defs and classes, so no side effects | 22:12 |
| *** dpawlik9 is now known as dpawlik | 22:13 | |
| openstackgerrit | Jeremy Stanley proposed opendev/gerritbot master: Don't reconnect on MessageTooLong exceptions https://review.opendev.org/c/opendev/gerritbot/+/781920 | 22:13 |
| *** Dmitrii-Sh has quit IRC | 22:13 | |
| *** janders has quit IRC | 22:13 | |
| *** Dmitrii-Sh8 is now known as Dmitrii-Sh | 22:13 | |
| *** brinzhang0 has quit IRC | 22:13 | |
| *** tobiash has quit IRC | 22:13 | |
| *** tobiash has joined #opendev | 22:14 | |
| *** mgoddard has quit IRC | 22:16 | |
| *** jpena|off has quit IRC | 22:16 | |
| *** gothicserpent has joined #opendev | 22:16 | |
| *** jpena|off has joined #opendev | 22:18 | |
| *** akahat has joined #opendev | 22:19 | |
| *** sshnaidm|off has quit IRC | 22:54 | |
| *** hamalq has quit IRC | 23:05 | |
| *** tkajinam has quit IRC | 23:27 | |
| *** mlavalle has quit IRC | 23:37 | |
| *** stand has quit IRC | 23:47 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!