| opendevreview | OpenStack Proposal Bot proposed openstack/kolla master: Updated ansible-lint in lint-requirements.txt https://review.opendev.org/c/openstack/kolla/+/1006878 | 02:23 |
|---|---|---|
| opendevreview | OpenStack Proposal Bot proposed openstack/kolla master: Updated from check-sources https://review.opendev.org/c/openstack/kolla/+/1006144 | 02:29 |
| opendevreview | OpenStack Proposal Bot proposed openstack/kolla-ansible master: Updated ansible-lint j2lint in lint-requirements.txt https://review.opendev.org/c/openstack/kolla-ansible/+/1006881 | 02:51 |
| mnasiadka | morning | 05:33 |
| opendevreview | Merged openstack/kolla master: Updated ansible-lint in lint-requirements.txt https://review.opendev.org/c/openstack/kolla/+/1006878 | 05:39 |
| opendevreview | Merged openstack/kolla-ansible master: Updated ansible-lint j2lint in lint-requirements.txt https://review.opendev.org/c/openstack/kolla-ansible/+/1006881 | 05:46 |
| fprzewozn | morning! | 07:24 |
| fprzewozn | mnasiadka we would have to split ports or use query analysis | 07:30 |
| fprzewozn | btw any one has any experience with running Noble containers (like for 2026.1) on Resolute on nodes? | 07:32 |
| mnasiadka | fprzewozn: something like https://review.opendev.org/c/openstack/kolla-ansible/+/1006469 (for proxysql) | 07:33 |
| fprzewozn | yeah something like that | 07:34 |
| mnasiadka | But first we need to get CI stable | 07:34 |
| mnasiadka | So I can finish working on that | 07:34 |
| fprzewozn | but underneath it's still single interface for further connection towards db itself | 07:35 |
| mnasiadka | Why would we need more interfaces? | 07:45 |
| mnasiadka | We just want to make the reads to go to all nodes, and writes to writer_hostgroup just like today? | 07:46 |
| fprzewozn | I'm not saying we need it :) | 07:56 |
| opendevreview | Will Szumski proposed openstack/kayobe master: CI: Use OpenDev Galaxy mirror https://review.opendev.org/c/openstack/kayobe/+/1006581 | 08:36 |
| fprzewozn | mikal please add Depends-On https://review.opendev.org/c/openstack/kolla-ansible/+/1005460 for all other changes in https://review.opendev.org/q/topic:"json-logging" | 08:37 |
| fprzewozn | there's no relationship set | 08:38 |
| fprzewozn | and all PRs are "active" | 08:38 |
| mikal | fprzewozn: this is confusing to me. Should the subsequent ones just be a chain after the first one which has the depends-on? I'm happy to add it to the others, just confused about why its needed. | 08:40 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Wait for fluentd to tail late log files. https://review.opendev.org/c/openstack/kolla-ansible/+/1001633 | 08:44 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Deploy Kerbside with Kolla-Ansible. https://review.opendev.org/c/openstack/kolla-ansible/+/976889 | 08:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Use a routable IP for qemu SPICE consoles. https://review.opendev.org/c/openstack/kolla-ansible/+/967801 | 08:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Add kerbside CI scenario jobs. https://review.opendev.org/c/openstack/kolla-ansible/+/988189 | 08:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Run the spice-direct tempest test in the kerbside scenario. https://review.opendev.org/c/openstack/kolla-ansible/+/988913 | 08:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Run a Kerbside-fronted SPICE tempest test in the kerbside scenario. https://review.opendev.org/c/openstack/kolla-ansible/+/989614 | 08:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Collect etcd logs with Fluentd. https://review.opendev.org/c/openstack/kolla-ansible/+/1005458 | 08:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Detect log levels in JSON formatted logs. https://review.opendev.org/c/openstack/kolla-ansible/+/1005459 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Add a fluentd path for JSON formatted OpenStack logs. https://review.opendev.org/c/openstack/kolla-ansible/+/1005460 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Let Keystone log JSON. https://review.opendev.org/c/openstack/kolla-ansible/+/1005461 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the core services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005462 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the compute services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005463 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the networking services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005464 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the storage services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005465 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the telemetry services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005466 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the remaining services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005467 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Exercise JSON logging in the telemetry scenario. https://review.opendev.org/c/openstack/kolla-ansible/+/1005468 | 08:46 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Document JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005469 | 08:46 |
| mikal | fprzewozn: done | 08:46 |
| fprzewozn | https://review.opendev.org/c/openstack/kolla-ansible/+/1005462 has https://review.opendev.org/c/openstack/kolla-ansible/+/1005458 as a dependency, shouldn't that be dependent on https://review.opendev.org/c/openstack/kolla-ansible/+/1005460 ? | 08:49 |
| PrzemekK | can i ask what give those changes in https://review.opendev.org/q/topic:%22json-logging%22 . We use opensearch with central logging for search for critical/error events and it working fine. | 08:58 |
| opendevreview | Will Szumski proposed openstack/kayobe master: Wait for firewalld to be ready before configuring https://review.opendev.org/c/openstack/kayobe/+/1006594 | 09:01 |
| opendevreview | Pierre Riteau proposed openstack/kolla-ansible master: Document letsencrypt with single external frontend https://review.opendev.org/c/openstack/kolla-ansible/+/1006909 | 09:06 |
| frickler | fyi got an appointment before the meeting, not sure whether I'll be back in time | 09:11 |
| mikal | fprzewozn: this is why I am confused by the request for a depends-on -- the JSON logging stuff was already in a series in a gerrit topic with the first patch in the series depends-on'ing the foundation it builds on. So I am not sure why you asked me to add the depends-on to the other patches in the series. They're already a chain. | 09:28 |
| mikal | PrzemekK: noting that you've left, there are two reasons I did the JSON logging patches. Kolla-Ansible has two components that _only_ log in JSON -- etcd and kerbside. At the moment we just don't collect etcd logs as a result, but I figured I'd just fix it. Overall the advantage of JSON logging is that it means you don't have to parse the log | 09:30 |
| mikal | messages to find their fields, they're already labelled for you. So it should be mean better / faster field detection and searching in opensearch. | 09:30 |
| opendevreview | Will Szumski proposed openstack/kayobe master: Encourage agents to identify themselves https://review.opendev.org/c/openstack/kayobe/+/1006569 | 09:37 |
| fprzewozn | mikal https://docs.opendev.org/opendev/infra-manual/latest/developers.html#multiple-changes IMO all "Convert .. to JSON logging" should Depend-On "Add a fluentd path for JSON formatted OpenStack logs" | 09:40 |
| mikal | fprzewozn: I think that's for tying multiple branches together -- like how the JSON patches are linked to the Kerbside stuff which ones first. Within a single branch gerrit has handled order a series of commits for a very long time. That's what the "relation chain" at the top right of a gerrit patch index page is showing you. | 09:43 |
| mikal | fprzewozn: like I'll change it, but I don't think its actually needed and I thought I'd already changed it to what you asked for. | 09:44 |
| mnasiadka | I don’t think we ever discussed going to JSON, so that work risks being not accepted? | 09:44 |
| mikal | mnasiadka: that's true. Its one of the reasons its a separate branch. Its also not a cutover thing, its an option people can turn on, but noting that etcd logging is currently not shipped. | 09:45 |
| mnasiadka | Well, not being a cutover thing means not being tested - with all of it’s outcomes (people being frustrated that it doesn’t work, raising bugs, we’re not good at maintaining bugs, etc) | 09:47 |
| mikal | One of the latter patches in the series converts an existing CI job to turning JSON on so that it can assert that logging works, so it would at least see some CI. | 09:48 |
| tafkamax | I think we should merge more and use more CI | 09:48 |
| opendevreview | Pierre Riteau proposed openstack/kayobe master: CI: Ignore .md files https://review.opendev.org/c/openstack/kayobe/+/1006914 | 09:49 |
| tafkamax | if option is false by default then oh well, maybe some CI job might hit it. | 09:49 |
| tafkamax | like my current usage with nova_safety_upgrade | 09:49 |
| opendevreview | Pierre Riteau proposed openstack/kayobe master: Encourage agents to identify themselves https://review.opendev.org/c/openstack/kayobe/+/1006569 | 09:50 |
| tafkamax | overall it is inevitable that errors come out, but missing entire logging for an component should not be okay... | 09:50 |
| mnasiadka | More CI means we use more of OpenDev’s resources (which we already use a lot) - nothing is made out of rubber | 09:50 |
| tafkamax | and maybe @ping people when they were the ones who introduced the patch that rboke stuff | 09:51 |
| mnasiadka | tafkamax: pinging people that are not here anymore doesn’t help :) | 09:51 |
| tafkamax | yeah, but I meant that in the current case the patches are from active contrributors | 09:51 |
| fprzewozn | yeahh to not put ourselves in https://review.opendev.org/c/openstack/kolla-ansible/+/917434 https://bugs.launchpad.net/kolla-ansible/+bug/2064104 and similiar again | 09:51 |
| opendevreview | Will Szumski proposed openstack/kayobe master: Fix Ironic inspection kernel options https://review.opendev.org/c/openstack/kayobe/+/1003856 | 09:52 |
| mikal | In this case its also not more CI jobs -- it recycles an existing job. | 09:52 |
| blanson[m] | frickler: https://review.opendev.org/c/openstack/kolla-ansible/+/988415/3 seems to be passing now, so hopefully that lifts your interrogations about https://review.opendev.org/c/openstack/kolla-ansible/+/986237/4 :) | 09:52 |
| mikal | I think its a bit unfair to imply that I will dump patches and run away. I've been an active OpenStack contributor since 2011 and have been working on the Kerbside stuff for at least two years. | 09:53 |
| tafkamax | mikal: that is good | 09:53 |
| tafkamax | I think if a CI job exists that is good and if it is resued even better. When introducing changes that don't introduce CI coverage, there should be warning in DOCS or somewhere. | 09:55 |
| tafkamax | E.g. if an option is default false and not included. | 09:56 |
| tafkamax | So bugs appear and we can link to the coverage not being there. | 09:57 |
| mnasiadka | Well, we had a discussion on switching all logs to journal, and this didn’t have a uniform acceptation - I just recommend people to discuss such things on the weekly meetings with a proposal or on the PTG | 09:59 |
| tafkamax | ok | 10:01 |
| mikal | mnasiadka: but honestly we discussed Kerbside at the last PTG and that hasn't translated into reviews. So surely a concrete proposal in the form of patches is relatively harmless? You don't have to merge them if you don't want to. | 10:01 |
| mnasiadka | I just don’t want you to be disappointed :) | 10:06 |
| blanson[m] | I'm kind of in favor of offering json logging options if we're having this conversation. we switched to it a while ago and it made our life so much easier. we swapped out fluentd for now from k-a cause it doesn't parse it but imo it'd be a good addition | 10:09 |
| mnasiadka | I think we should have a discussion on the monitoring part as a whole | 10:11 |
| mnasiadka | monitoring/logging | 10:11 |
| blanson[m] | yep | 10:12 |
| blanson[m] | do you have topics in particular that you'd like to discuss ? I can do a little list of them so that we can discuss it during weekly meet | 10:13 |
| mikal | Going back a conversation, I am still unclear about why the existing gerrit relation chain is insufficient for a series of patches? I'll put Depends-On in every subsequent patch in the series if I really need to, but I am confused about why its needed. | 10:14 |
| blanson[m] | mikal: chain dependency should suffice for patches in the same repo. that's what I was told and that's what I've always done and it never caused issue ? | 10:15 |
| blanson[m] | I'm a gerrit noob but Depend-On is for cross repo stuff right ? | 10:15 |
| fprzewozn | Yeah maybe I'm also misunderstanding the developers guide | 10:15 |
| mnasiadka | Depends-on for cross repo, right | 10:15 |
| mnasiadka | If it works if bottom patch in the chain has it (and I assume it should) - then it’s fine | 10:16 |
| opendevreview | Pierre Riteau proposed openstack/kayobe master: Remove outdated core reviewer list https://review.opendev.org/c/openstack/kayobe/+/1006924 | 10:16 |
| blanson[m] | fprzewozn: chained patches are rebased on top on each other, so if they're all in the same repo, they do not need Depends-On. but if you patch Kolla and K-A, then you add Depends-On because those are completely decoupled | 10:16 |
| blanson[m] | where did you get the misunderstanding from in the dev guides ? maybe we can improve it | 10:17 |
| mikal | mnasiadka: Depends-On is for cross _branch_ dependencies, including branches in different repos. | 10:17 |
| mnasiadka | Well, cross branch dependencies in the same repo are unusual | 10:18 |
| mikal | A stack of patches in a single branch doesn't need that because it uses the gerrit relations thing instead. The relations thing predates Depends-On, which was added to handle the case where you were depending on a change in something like Oslo. | 10:18 |
| fprzewozn | okey, my bad | 10:18 |
| fprzewozn | sorry for the noice then | 10:19 |
| mikal | No its fine, I was just confused because I couldn't figure out what I was missing. I'll remove them again when I push the reply to your 976889 comments. | 10:19 |
| fprzewozn | okey! | 10:23 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Deploy Kerbside with Kolla-Ansible. https://review.opendev.org/c/openstack/kolla-ansible/+/976889 | 10:38 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Use a routable IP for qemu SPICE consoles. https://review.opendev.org/c/openstack/kolla-ansible/+/967801 | 10:38 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Add kerbside CI scenario jobs. https://review.opendev.org/c/openstack/kolla-ansible/+/988189 | 10:38 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Run the spice-direct tempest test in the kerbside scenario. https://review.opendev.org/c/openstack/kolla-ansible/+/988913 | 10:38 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Run a Kerbside-fronted SPICE tempest test in the kerbside scenario. https://review.opendev.org/c/openstack/kolla-ansible/+/989614 | 10:38 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Collect etcd logs with Fluentd. https://review.opendev.org/c/openstack/kolla-ansible/+/1005458 | 10:38 |
| mikal | fprzewozn: thanks for your detailed review on 976889. This is updates to address your comments ---^ | 10:38 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Detect log levels in JSON formatted logs. https://review.opendev.org/c/openstack/kolla-ansible/+/1005459 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Add a fluentd path for JSON formatted OpenStack logs. https://review.opendev.org/c/openstack/kolla-ansible/+/1005460 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Let Keystone log JSON. https://review.opendev.org/c/openstack/kolla-ansible/+/1005461 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the core services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005462 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the compute services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005463 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the networking services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005464 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the storage services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005465 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the telemetry services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005466 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Convert the remaining services to JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005467 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Exercise JSON logging in the telemetry scenario. https://review.opendev.org/c/openstack/kolla-ansible/+/1005468 | 10:45 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Document JSON logging. https://review.opendev.org/c/openstack/kolla-ansible/+/1005469 | 10:45 |
| mikal | ^---- reverts the depends-on thing. | 10:46 |
| PrzemekK | @mikal After configure use_json in nova there are no logs in opensearch (central logging) so also this part should be included from what i can see current parser works same/very simillar as logs in json https://paste.openstack.org/show/b5HQeoP2nVwAtn0L4s6w/ | 10:53 |
| mikal | Yes, that's what that JSON logging series adds -- support for oslo.logging's pre-existing JSON logging output option. It has the side effect that etcd and Kerbside logs can also be collected, but oslo.logging has supported JSON formatted output for quite some time. | 10:55 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Deploy Kerbside with Kolla-Ansible. https://review.opendev.org/c/openstack/kolla-ansible/+/976889 | 11:18 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Use a routable IP for qemu SPICE consoles. https://review.opendev.org/c/openstack/kolla-ansible/+/967801 | 11:18 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Add kerbside CI scenario jobs. https://review.opendev.org/c/openstack/kolla-ansible/+/988189 | 11:18 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Run the spice-direct tempest test in the kerbside scenario. https://review.opendev.org/c/openstack/kolla-ansible/+/988913 | 11:18 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Run a Kerbside-fronted SPICE tempest test in the kerbside scenario. https://review.opendev.org/c/openstack/kolla-ansible/+/989614 | 11:18 |
| opendevreview | Michael Still proposed openstack/kolla-ansible master: Collect etcd logs with Fluentd. https://review.opendev.org/c/openstack/kolla-ansible/+/1005458 | 11:18 |
| mikal | fprzewozn: further replies to your comments ----^ | 11:18 |
| mikal | fprzewozn: it is however now past my bedtime, so I will check in tomorrow morning and see if you have any other questions. Thanks again for your reviews. | 11:19 |
| opendevreview | Merged openstack/kolla-ansible stable/2026.1: kolla-toolbox: remove leftover kolla-toolbox.json.j2 https://review.opendev.org/c/openstack/kolla-ansible/+/997381 | 12:22 |
| mnasiadka | bbezak frickler kevko mmalchuk gkoper jovial mattcrees dougszu darmach pabloclsn ravlew salmankh amir58118 r-krcek blanson[m] fprzewozn AlmaMC[m] - meeting in 10 minutes | 12:50 |
| fprzewozn | having 8 mins before meeting I'll ask one more time now: anyone has any experience with running Noble containers (like for 2026.1) on Resolute on nodes? | 12:52 |
| priteau | fprzewozn: this is currently forbidden by host OS prechecks, so I doubt it has been done much | 12:58 |
| priteau | I believe support for 26.04 is planned for the next cycle | 12:58 |
| fprzewozn | priteau you mean in 2026.2? | 13:00 |
| priteau | yes | 13:00 |
| priteau | No sorry, 2027.1 | 13:00 |
| priteau | With some backports to 2026.1 for compatibility | 13:00 |
| mnasiadka | 2026.2 PTI does not include testing on Resolute, so I don’t think we want to go crazy | 13:00 |
| mnasiadka | #startmeeting kolla | 13:01 |
| opendevmeet | Meeting started Wed Sep 23 13:01:05 2026 UTC and is due to finish in 60 minutes. The chair is mnasiadka. Information about MeetBot at http://wiki.debian.org/MeetBot. | 13:01 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 13:01 |
| opendevmeet | The meeting name has been set to 'kolla' | 13:01 |
| ravlew | o/ | 13:01 |
| mnasiadka | #topic rollcall | 13:01 |
| mnasiadka | O/ | 13:01 |
| dougszu | o/ | 13:01 |
| MurrayGroves | o/ | 13:01 |
| bbezak | o// | 13:01 |
| priteau | o/ | 13:01 |
| blanson[m] | hello ! | 13:01 |
| fprzewozn | o7 | 13:01 |
| mnasiadka | #topic agenda | 13:02 |
| mnasiadka | * CI status | 13:02 |
| mnasiadka | * Release tasks | 13:02 |
| mnasiadka | * Open discussion | 13:03 |
| mnasiadka | #topic CI status | 13:03 |
| mnasiadka | So, CI has been not stable recently, mainly due to VexxHost ipv6 problems - it’s getting a bit better - but still not resolved from OpenDev infra side | 13:03 |
| mnasiadka | #topic Release tasks | 13:03 |
| mnasiadka | That’s the main topic of the day, Hibiscus (2026.2) is getting released next week - and this time we’re trying to stick to branching and tagging rc1 in nearly the same timeframe | 13:04 |
| mnasiadka | First Kolla, then Kolla-Ansible and a-c-k, Kayobe as the last one | 13:04 |
| mnasiadka | Tenks (being our new deliverable) doesn’t seem to be branching and tagging | 13:05 |
| mnasiadka | Any questions? Proposals? Patches that HAVE to be merged? | 13:06 |
| fprzewozn | so, we're waiting with features now? | 13:06 |
| mertongngl | Hi there, https://review.opendev.org/c/openstack/kolla-ansible/+/998499 is there any blocker to take this development to 2026.1's backport? | 13:06 |
| opendevreview | Merged openstack/kolla-ansible master: Fix ETCD health check hardcoded to HTTP https://review.opendev.org/c/openstack/kolla-ansible/+/1005538 | 13:07 |
| mnasiadka | fprzewozn: yes, according to the kolla release guide we’re in feature freeze | 13:08 |
| fprzewozn | o7 | 13:09 |
| mnasiadka | mertongngl: no, that’s a bug - feel free to raise cherry-picks to stable branches | 13:09 |
| blanson[m] | we have a bunch of features in RP that are waiting for the feature freeze to end in k-a, can we mark stuff up in the coming week(s) so that it doesn't drag on for another 4 months to be merged ? I'm thinking mainly about vault barbican integration, keystone credential encryption and some other that are up there. | 13:11 |
| fprzewozn | got few others myself :) | 13:12 |
| mnasiadka | Only 4? | 13:12 |
| blanson[m] | most of those have been reviewed but merging them now is not really a good time | 13:12 |
| blanson[m] | I mean I can make a list | 13:12 |
| mnasiadka | I think we need to try to manage RP+1 in a better way | 13:13 |
| blanson[m] | for stuff "approved but waiting on FF to end" | 13:13 |
| opendevreview | Merged openstack/kayobe master: CI: Ignore .md files https://review.opendev.org/c/openstack/kayobe/+/1006914 | 13:13 |
| opendevreview | Merged openstack/kayobe master: Encourage agents to identify themselves https://review.opendev.org/c/openstack/kayobe/+/1006569 | 13:13 |
| mnasiadka | But maybe it’s not the time to discuss it yet | 13:13 |
| blanson[m] | agreed, tho I have no idea how | 13:13 |
| blanson[m] | either way, I'll make a list of features un RP that seem fine, but are just dragging on | 13:14 |
| mnasiadka | Weekly prios which we will ignore, RP+1 which we just pile on, festivals of reviews which don’t happen - choose one :) | 13:14 |
| blanson[m] | so we have a bunch of stuff to merge hopefully for 2027.1 | 13:14 |
| mnasiadka | Yeah, I agree - let’s try to organise some ,,maintainer meeting’’ after we get final 2026.2 tagged to think how to manage it better | 13:15 |
| blanson[m] | cool I'll work on that | 13:15 |
| mnasiadka | Ok then, let’s move to open discussion | 13:15 |
| blanson[m] | for 2026.2 tho I think there isn't many more bugs/blockers to merge ? | 13:15 |
| mnasiadka | I think we merged the main things that were required | 13:16 |
| mnasiadka | I’ll run some checks to make sure we’re good before branching | 13:16 |
| mnasiadka | Ok then, open discussion | 13:16 |
| mnasiadka | #topic Open discussion | 13:16 |
| mnasiadka | Anybody anything? | 13:16 |
| MurrayGroves | For https://review.opendev.org/c/openstack/kolla/+/987109, basing the image on Rocky is quite involved, since Rocky 10 dropped Xorg in favour of Wayland, and there are packages missing/outdated. Current plan is to go ahead with Ubuntu and Debian base images for now, with support for Rocky coming later. Does this seem ok to everyone? | 13:17 |
| mnasiadka | I think this is fine as an interim solution, but do you expect Kolla-Ansible to fail on prechecks when somebody wants to deploy this on Rocky? Or what is the plan? | 13:18 |
| dougszu | We could offer the Ubuntu image as an artefact a bit like IPA to begin with? | 13:19 |
| mnasiadka | So deploy Ubuntu based image on Rocky instead of failing | 13:19 |
| mnasiadka | blanson[m], bbezak, others ^^ opinions? | 13:20 |
| blanson[m] | I like dougszu 's idea | 13:20 |
| blanson[m] | tho it requires special care cause the container needs to work on both host system | 13:20 |
| opendevreview | Pierre Riteau proposed openstack/kayobe stable/2026.1: CI: Ignore .md files https://review.opendev.org/c/openstack/kayobe/+/1006945 | 13:21 |
| mnasiadka | Yeah, I’ve seen kolla-toolbox from Ubuntu failing on Rocky in the past | 13:21 |
| blanson[m] | it doesn't really conform to the design philosophy of the project, but the alternative it probbly to have a really bad time making it work on Rocky according to MurrayGroves so | 13:22 |
| mnasiadka | So hopefully that works as an experimental/interim solution | 13:22 |
| MurrayGroves | Getting it working on Rocky would involve building from source somewhere, or getting a package version bumped in EPEL | 13:22 |
| dougszu | hopefully longer term we can work on the latter | 13:23 |
| mnasiadka | I think worst case scenario we can use openstack.kolla org in Fedora COPR to get that built - or just work on a Wayland alternative | 13:23 |
| MurrayGroves | Plain Fedora repos have the correct version since a couple months ago so hopefully EPEL will catch up | 13:23 |
| MurrayGroves | This is with Wayland alternative mnasiadka | 13:24 |
| MurrayGroves | Trying to do Xorg would be worse because we would need to build Xorg | 13:24 |
| mnasiadka | Epel 10.4 doesn’t have them yet? | 13:25 |
| MurrayGroves | Them being what sorry | 13:25 |
| mnasiadka | You said that Fedora repos have correct versions - so I’m asking if they are being updated as part of EPEL 10.4, or not yet ;-) | 13:25 |
| MurrayGroves | 10.4 still has incorrect version | 13:26 |
| MurrayGroves | https://packages.fedoraproject.org/pkgs/cage/cage/index.html | 13:26 |
| MurrayGroves | We need 0.3.x | 13:26 |
| mnasiadka | Ok then - it might be worth raising a bugzilla EPEL request if that doesn’t get updated soon | 13:29 |
| MurrayGroves | Ok makes sense, how long is "soon" in package maintainer land? | 13:30 |
| mnasiadka | Have no clue - maybe https://docs.fedoraproject.org/en-US/epel/epel-package-request/ has some hints | 13:31 |
| MurrayGroves | Will investigate | 13:31 |
| mnasiadka | Ok then, any other topic? | 13:31 |
| mnasiadka | Anybody needs an FFE? (Feature Freeze Exception) | 13:31 |
| mnasiadka | Guess not | 13:34 |
| mnasiadka | Giving you back some time :) | 13:34 |
| mnasiadka | Thank you all for coming | 13:34 |
| mnasiadka | #endmeeting | 13:34 |
| opendevmeet | Meeting ended Wed Sep 23 13:34:47 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 13:34 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/kolla/2026/kolla.2026-09-23-13.01.html | 13:34 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/kolla/2026/kolla.2026-09-23-13.01.txt | 13:34 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/kolla/2026/kolla.2026-09-23-13.01.log.html | 13:34 |
| blanson[m] | thanks | 13:34 |
| MurrayGroves | Thank you! | 13:35 |
| ravlew | thanks | 13:35 |
| blanson[m] | mnasiadka: I'll move the feature patches to feature freeze so hopefully the RP panel is not as intimidating and we can think of some better way to manage this whole thing | 13:38 |
| dougszu | thanks | 13:40 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!