| opendevreview | Verification of a change to openstack/kolla stable/2025.1 failed: [2025.1] heat: pin setuptools below 82 https://review.opendev.org/c/openstack/kolla/+/978553 | 00:42 |
|---|---|---|
| opendevreview | Verification of a change to openstack/kolla stable/2025.1 failed: [2025.1] CI: enable k-a rocky10 https://review.opendev.org/c/openstack/kolla/+/978729 | 00:42 |
| opendevreview | Verification of a change to openstack/kolla-ansible stable/2025.1 failed: [2025.1] pin setuptools in tox docs https://review.opendev.org/c/openstack/kolla-ansible/+/977959 | 00:42 |
| opendevreview | Pierre Riteau proposed openstack/kayobe master: [networkd] Fix broken conditional https://review.opendev.org/c/openstack/kayobe/+/975173 | 07:28 |
| opendevreview | Pierre Riteau proposed openstack/kayobe master: [firewalld] Fix broken conditional https://review.opendev.org/c/openstack/kayobe/+/975186 | 07:29 |
| opendevreview | Pierre Riteau proposed openstack/kayobe master: Revert "CI: Disable seed jobs" https://review.opendev.org/c/openstack/kayobe/+/978855 | 07:33 |
| mnasiadka | Morning | 07:36 |
| blanson[m] | hello ! | 08:12 |
| blanson[m] | :) https://review.opendev.org/q/owner:bertrand.lanson@infomaniak.com+-branch:+master+repo:openstack/kolla-ansible I have some backports that have been waiting for a while if anyone can have a look :) | 08:12 |
| thelounge87 | Hi | 08:22 |
| thelounge87 | So when exactly is this going to happen? the meeting | 08:22 |
| opendevreview | Leonie Chamberlin-Medd proposed openstack/kayobe master: Add support for fail2ban in Kayobe https://review.opendev.org/c/openstack/kayobe/+/973090 | 08:33 |
| *** thelounge870 is now known as thelounge87 | 08:34 | |
| thelounge87 | Hi Hi HI | 08:35 |
| mnasiadka | thelounge87: https://meetings.opendev.org/#Kolla_Team_Meeting | 08:35 |
| thelounge87 | can anyone help me with sso | 08:35 |
| thelounge87 | I am encountering a weird error, where the logs dont show anything sus. But the horizon dashboard after succesful OIDC completion shows | 08:36 |
| thelounge87 | ```Login failed: An error occurred authenticating. Please try again later.``` | 08:36 |
| thelounge87 | But the keystone logs are perfectly fine. Even in debug mode, it showed a neat empheral user created successfully etc | 08:37 |
| thelounge87 | My globals.yml relating to that is: | 08:38 |
| thelounge87 | ``` | 08:38 |
| thelounge87 | horizon_keystone_domain_choices: | 08:38 |
| thelounge87 | Default: Default | 08:38 |
| thelounge87 | azure_sso: azure_sso | 08:38 |
| thelounge87 | horizon_keystone_multidomain: true | 08:38 |
| thelounge87 | keystone_federation_oidc_additional_options: | 08:38 |
| thelounge87 | OIDCClaimPrefix: OIDC- | 08:38 |
| thelounge87 | OIDCPassIDTokenAs: claims | 08:38 |
| thelounge87 | OIDCResponseType: code | 08:38 |
| thelounge87 | OIDCXForwardedHeaders: X-Forwarded-Proto | 08:38 |
| thelounge87 | keystone_federation_oidc_jwks_uri: https://login.microsoftonline.com/e54d02ca-662d-401b-9995-ee65d91d85c7/discovery/v2.0/keys | 08:38 |
| thelounge87 | keystone_identity_mappings: | 08:38 |
| thelounge87 | - file: /etc/kolla/config/keystone/federation/oidc/attribute_maps/azure_ad_mapping.json | 08:38 |
| thelounge87 | name: azure_ad_mapping | 08:38 |
| thelounge87 | keystone_identity_providers: | 08:38 |
| thelounge87 | - attribute_mapping: azure_ad_mapping | 08:38 |
| thelounge87 | identifier: https://login.microsoftonline.com/e54d02ca-662d-401b-9995-ee65d91d85c7/v2.0 | 08:38 |
| thelounge87 | metadata_folder: /etc/kolla/config/keystone/federation/oidc/metadata | 08:38 |
| thelounge87 | name: azure-ad | 08:38 |
| thelounge87 | openstack_domain: azure_sso | 08:38 |
| thelounge87 | protocol: openid | 08:38 |
| thelounge87 | public_name: Authenticate via Azure AD | 08:38 |
| thelounge87 | ``` | 08:38 |
| thelounge87 | sorry I am new to IRC so (not sure how to show attachments or long text) | 08:39 |
| tafkamax | What release are you using | 08:48 |
| tafkamax | what do the console logs in browser show though? | 08:49 |
| tafkamax | I havent't used the AD oidc myself. Do you have the mapping.json correct? | 08:49 |
| thelounge87 | Using 2025.1 . Yes I tried many mappings and edited and all result in the same result. | 08:50 |
| thelounge87 | tafkamaxVishwa Mithra should I use pastebin to show rather than in chat? | 08:51 |
| thelounge87 | ``` | 08:52 |
| thelounge87 | (apples) root@kolla1:/home/ubuntu# cat /etc/kolla/config/keystone/federation/oidc/attribute_maps/azure_ad_mapping.json | 08:52 |
| thelounge87 | [ | 08:52 |
| thelounge87 | { | 08:52 |
| thelounge87 | "local": [ | 08:52 |
| thelounge87 | { | 08:52 |
| thelounge87 | "user": { | 08:52 |
| thelounge87 | "name": "{0}", | 08:52 |
| thelounge87 | "email": "{1}", | 08:52 |
| thelounge87 | "type": "ephemeral" | 08:52 |
| thelounge87 | }, | 08:52 |
| thelounge87 | "group": { | 08:52 |
| thelounge87 | "name": "cloud-users", | 08:52 |
| thelounge87 | "domain": { | 08:52 |
| thelounge87 | "name": "azure_sso" | 08:52 |
| thelounge87 | } | 08:52 |
| thelounge87 | } | 08:52 |
| thelounge87 | } | 08:52 |
| tafkamax | hmm horizon_keystone_domain_choices. I haven't used that option. | 08:52 |
| thelounge87 | ], | 08:52 |
| thelounge87 | "remote": [ | 08:52 |
| thelounge87 | { "type": "OIDC-email" }, | 08:52 |
| thelounge87 | { "type": "OIDC-email" } | 08:52 |
| thelounge87 | ] | 08:52 |
| thelounge87 | } | 08:52 |
| thelounge87 | ]``` | 08:52 |
| tafkamax | pastebin would be better | 08:52 |
| tafkamax | For us we use a single domain, but with multiple authentication methods. | 08:53 |
| thelounge87 | mmmm so domain could be an issue? | 08:53 |
| tafkamax | I am not sure, I havent' tried it like so. | 08:54 |
| thelounge87 | ``` | 08:54 |
| thelounge87 | 2026-03-03 11:10:17.715 1081 INFO keystone.identity.core [None req-ceebc006-2c74-46e0-800e-f04eb28a95d7 - - - - - -] Adding user [{'email': 'testuser1@Bdxworld.com', 'id': 'b3e67718c6187dbcab031b249eb72133cb54614fea325cbd8e58650e9947fbf4', 'name': 'testuser1@galamTech.com', 'domain_id': '6bbcdd1d154844b3963e5dc6be75462d', 'enabled': True, | 08:54 |
| thelounge87 | 'password_expires_at': None, 'options': {}}] to group [64fc4d5fca1643f48df24f5c6a437c0c].``` | 08:54 |
| thelounge87 | ignore the bdxworld and galamTech lol | 08:54 |
| tafkamax | The metadata folder has all of the specified components that is in the docs? | 08:55 |
| thelounge87 | Yes | 08:55 |
| thelounge87 | ``` | 08:55 |
| thelounge87 | (apples) root@kolla1:/home/ubuntu# ls /etc/kolla/config/keystone/federation/oidc/ | 08:55 |
| thelounge87 | attribute_maps cert metadata | 08:55 |
| thelounge87 | (apples) root@kolla1:/home/ubuntu# ls /etc/kolla/config/keystone/federation/oidc/metadata/ | 08:55 |
| thelounge87 | login.microsoftonline.com%2Fe54d02ca-662d-401b-9995-ee65d91d85c7%2Fv2.0.client | 08:55 |
| thelounge87 | login.microsoftonline.com%2Fe54d02ca-662d-401b-9995-ee65d91d85c7%2Fv2.0.conf | 08:55 |
| thelounge87 | login.microsoftonline.com%2Fe54d02ca-662d-4``` | 08:55 |
| tafkamax | I can't spot anything obvious. | 08:56 |
| thelounge87 | mmmm, yeah, I'll try removing the domain and try | 08:56 |
| tafkamax | The only thing different for me is that I have a single domain but multiple authentication methods. | 08:56 |
| tafkamax | And I don't specify the jwks uri. And also the additional_options. | 08:57 |
| thelounge87 | those were added after checking the logs tbh. It asked that the proto forwarding doesnt exist so added that | 08:57 |
| tafkamax | ok | 08:58 |
| opendevreview | Merged openstack/kolla-ansible stable/2025.2: Fix trove_api healthcheck with TLS backend https://review.opendev.org/c/openstack/kolla-ansible/+/972520 | 09:02 |
| *** EugenMayer4401802 is now known as EugenMayer440180 | 09:08 | |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: glance: Add configuration for cinder backend https://review.opendev.org/c/openstack/kolla-ansible/+/978863 | 09:31 |
| opendevreview | Verification of a change to openstack/kolla-ansible stable/2025.2 failed: Fix variable name typo in haproxy single external frontend https://review.opendev.org/c/openstack/kolla-ansible/+/971594 | 09:32 |
| opendevreview | Doug Szumski proposed openstack/kolla-ansible master: Restart only changed iterable containers https://review.opendev.org/c/openstack/kolla-ansible/+/976659 | 09:40 |
| *** jhorstmann is now known as Guest4298 | 10:00 | |
| opendevreview | Merged openstack/kolla-ansible stable/2025.2: Fix idempotence on comparing capabilities for podman https://review.opendev.org/c/openstack/kolla-ansible/+/972830 | 10:03 |
| opendevreview | Merged openstack/kolla-ansible stable/2025.2: Fix podman idempotence on comparing container dimensions https://review.opendev.org/c/openstack/kolla-ansible/+/972831 | 10:03 |
| opendevreview | Verification of a change to openstack/kolla-ansible stable/2025.2 failed: Fix idempotence on podman volume comparison https://review.opendev.org/c/openstack/kolla-ansible/+/972832 | 10:10 |
| opendevreview | Verification of a change to openstack/kolla-ansible stable/2025.2 failed: Test idempotence of reconfigure command in CI https://review.opendev.org/c/openstack/kolla-ansible/+/972833 | 10:10 |
| opendevreview | Seunghun Lee proposed openstack/kolla-ansible master: Improve ProxySQL test https://review.opendev.org/c/openstack/kolla-ansible/+/978221 | 10:22 |
| opendevreview | Merged openstack/kayobe master: Revert "CI: Disable seed jobs" https://review.opendev.org/c/openstack/kayobe/+/978855 | 10:32 |
| thelounge87 | Did anyone get issues running 2025.2? | 10:49 |
| thelounge87 | The haproxy keep alived is not generating the haproxy backend config | 10:50 |
| tafkamax | I am running 2025.2, have not seen the issue you mentioned. | 10:51 |
| thelounge87 | mmmmm | 10:51 |
| tafkamax | But for IDP config you need to currently run this fix https://bugs.launchpad.net/kolla-ansible/+bug/2134455/comments/1 | 10:51 |
| thelounge87 | I am trying to upgrade from 2025.1 to 2025.2 | 10:51 |
| tafkamax | I haven't tried the 2025.1 to 2025.2 upgrade path yet. | 10:52 |
| tafkamax | I have done a fresh install of 2025.2 so far. | 10:52 |
| thelounge87 | Yeah, I am trying to upgrade and it keeps failing with a myriad of issues | 10:54 |
| opendevreview | Seunghun Lee proposed openstack/kolla-ansible master: Improve ProxySQL test https://review.opendev.org/c/openstack/kolla-ansible/+/978221 | 10:54 |
| thelounge87 | I have one question who tf introduced proxySQL? | 10:58 |
| thelounge87 | why do we even need it?? | 10:58 |
| thelounge87 | protect the service from ohmysql client? or sqlalchemy? or what? if a service doesn;t have access to the db , it will fail silently if proxysql is there. Else it will stop and throw an error. | 10:58 |
| thelounge87 | throwing an error is any day better than being silent | 10:59 |
| tafkamax | it is there for better HA capabilities | 10:59 |
| tafkamax | I guess you can look in the code PR-s where it was introduced, why it was introduced. | 10:59 |
| tafkamax | I think before the haproxy -> mariadb had it's issues. | 11:00 |
| tafkamax | Quick AI query might give you an approximate answer :D | 11:01 |
| opendevreview | Pierre Riteau proposed openstack/ansible-collection-kolla master: Fix check for Docker containerd storage https://review.opendev.org/c/openstack/ansible-collection-kolla/+/978172 | 11:08 |
| opendevreview | Pierre Riteau proposed openstack/kolla-ansible master: [DNM] Test a-c-k fix https://review.opendev.org/c/openstack/kolla-ansible/+/978178 | 11:11 |
| opendevreview | Pierre Riteau proposed openstack/kolla-ansible master: Revert "CI: Pin docker to <29" https://review.opendev.org/c/openstack/kolla-ansible/+/978170 | 11:11 |
| thelounge87 | idk tbh, cause VIP and haproxy solves al the issues anyway. It seems more like extra clunk | 11:12 |
| thelounge87 | and silent failing is far more dangerous | 11:12 |
| thelounge87 | Also asked AI. | 11:13 |
| thelounge87 | and it mirrors my opinon due to psychofancy | 11:13 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: neutron: Switch neutron_plugin_agent default to ovn https://review.opendev.org/c/openstack/kolla-ansible/+/970114 | 11:13 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: neutron: Switch neutron_plugin_agent default to ovn https://review.opendev.org/c/openstack/kolla-ansible/+/970114 | 11:15 |
| mnasiadka | thelounge87: Have you come here to express your worries about the state of the project and willingness to pick up maintenance - or you just wanted to complain and lay out your frustration? | 11:22 |
| thelounge87 | mnasiadkaVishwa Mithra I actually have a multiregion on which I think you commented on so, yeah. Wanted to talk about that | 11:23 |
| thelounge87 | and yep, also want to contribute to the fixes too! | 11:24 |
| thelounge87 | Again, new to this, so not sure, how to go about it. (Is complaining bad or seen as such here?) | 11:24 |
| opendevreview | Pavel Shapovalov proposed openstack/kolla-ansible master: destroy: fix Octavia interface cleanup permissions https://review.opendev.org/c/openstack/kolla-ansible/+/978869 | 11:24 |
| thelounge87 | mnasiadkaVishwa Mithra https://review.opendev.org/c/openstack/kolla-ansible/+/977760 | 11:26 |
| thelounge87 | This one :) | 11:26 |
| tafkamax | I guess it is important to understand why some decisions were made. I remember 2024.2 or 2024.1 release without proxysql. The haproxy sent traffic to read-only mariadb nodes. With the migration to proxysql the node which allows writes gets it every time, thanks to proxysql being "smarter" than haproxy. | 11:27 |
| tafkamax | Since the usage of proxysql the mariadb usage has become more stable for us. | 11:27 |
| thelounge87 | And thats the reason deploymnets without proxysql in 2025.1 trying to upgrade to 2025.2 is breaking? | 11:29 |
| blanson[m] | quick question for people familiar with ironic. I'm reading about the novnc-proxy requirement, but: is it still required, and can kolla deploy it ? I can't find anything about it apart from "it's rerquired" in the ironic doc | 11:29 |
| tafkamax | havnet used ironic | 11:32 |
| tafkamax | thelounge87: Are you not using proxysql? | 11:33 |
| tafkamax | Because in the docs there is this important part: https://docs.openstack.org/releasenotes/kolla-ansible/2025.2.html#relnotes-21-0-0-stable-2025-2 | 11:33 |
| thelounge87 | No I am not | 11:33 |
| tafkamax | It is an interesting path. We migrated to proxysql before the 2025.2 upgrade. As some of these things are done before fully removal/deprecation. | 11:34 |
| opendevreview | Pavel Shapovalov proposed openstack/kolla-ansible master: destroy: fix Octavia interface cleanup permissions https://review.opendev.org/c/openstack/kolla-ansible/+/978869 | 11:34 |
| thelounge87 | ....why its now considered legacy? | 11:35 |
| thelounge87 | https://review.opendev.org/c/openstack/kolla-ansible/+/962479 Like...who takes these decisions? HP ? | 11:35 |
| thelounge87 | What about deployments who used nomral HAproxy and VIP and it works awesome as is | 11:35 |
| tafkamax | Because it had its problems and proxysql solved those. I guess the burden of maintaing two different LB-s for mariadb is not useful in this case, when one is better than the other. | 11:36 |
| thelounge87 | so make a breaking change and just drop it? instead of making a clear documentation reference as to how to upgrade if you do not have that enabled? | 11:37 |
| thelounge87 | sorry for getting frustated. | 11:38 |
| tafkamax | I think you need to check the release notes. It says that it has been removed. | 11:38 |
| tafkamax | In 2024.2 there is release notes that proxysql is default and you need to explicilty set it to false. | 11:39 |
| tafkamax | s/./ if you wish to use haproxy/ | 11:39 |
| thelounge87 | So im gonna revert back to 2025.1 -> enable proxysql -> upgrade kolla to 2025.2 and then try | 11:47 |
| mnasiadka | The upgrade process should be working, irrespective of the fact what you have configured before the upgrade - have you raised a bug? | 11:49 |
| tafkamax | I suppose if the enable_proxysql variable is false when upgrading to 2025.1 -> 2025.2 then it should fail? | 11:50 |
| tafkamax | I don't know the status of globals.yml | 11:50 |
| thelounge87 | it is failing when proxysql is set to false. I had not raised a bug. I am still trying to make it work. | 11:51 |
| opendevreview | Merged openstack/kolla stable/2025.1: [2025.1] heat: pin setuptools below 82 https://review.opendev.org/c/openstack/kolla/+/978553 | 11:51 |
| opendevreview | Merged openstack/kolla stable/2025.1: [2025.1] CI: enable k-a rocky10 https://review.opendev.org/c/openstack/kolla/+/978729 | 11:51 |
| thelounge87 | Assuming my globals.yml are wrong (And you need to have it correct) | 11:51 |
| thelounge87 | for the upgrade to work and proxysql to be there. | 11:51 |
| thelounge87 | One thing is clear: The globals.yml file which worked for 2025.1 does not work for 2025.2 as is. So yep. | 11:53 |
| thelounge87 | Not sure if its a bug or intentional. | 11:54 |
| thelounge87 | but could have been nice if mentioned in the upgrade guide. | 11:54 |
| opendevreview | Seunghun Lee proposed openstack/kolla-ansible master: Improve ProxySQL test https://review.opendev.org/c/openstack/kolla-ansible/+/978221 | 11:56 |
| mnasiadka | thelounge87: the expectation is that you’re reading release notes | 12:10 |
| thelounge87 | I think the general path for most who have legacy deployments is to just reuse the globals and compare it with new globals.yml and mix and match. But yep, maybe some kinda red highlighting would help instead of silent expectation :) similar to other docs. | 12:16 |
| thelounge87 | Yep that seems to do it for now. Note: ProxySQL is a must if you wanna upgrade from 2025.1 to 2025.2 (It breaks otherwise - especially if you specifically enable it as no in your globals.yml | 12:28 |
| thelounge87 | Doesn;t work new error. Don't do it. | 12:36 |
| thelounge87 | Maybe I am in the wrong here. But my strong opinon is that proxySQL is bad. And its a movement in the wrong direction for kolla ansible. Increasing complexity. mimicing worst practices of charmed openstack. | 12:39 |
| thelounge87 | just because canonical did it doesnt mean everyone should. | 12:39 |
| thelounge87 | And as said already, proxySQL fails silently because of really bad architecture on its end. And doesn't show up in either the service logs or db logs and has its own logs. Just keep increasing complexity . | 12:45 |
| thelounge87 | Can I talk to this guy Seughun Lee ? | 12:45 |
| opendevreview | Merged openstack/kolla-ansible stable/2025.1: [2025.1] pin setuptools in tox docs https://review.opendev.org/c/openstack/kolla-ansible/+/977959 | 12:51 |
| opendevreview | Merged openstack/kolla-ansible stable/2025.1: requirements: pin bcrypt to supported version https://review.opendev.org/c/openstack/kolla-ansible/+/976104 | 13:01 |
| opendevreview | Seunghun Lee proposed openstack/kolla-ansible master: Improve ProxySQL test https://review.opendev.org/c/openstack/kolla-ansible/+/978221 | 13:05 |
| opendevreview | Merged openstack/kolla stable/2025.1: debian: use valkey from bookworm-backports https://review.opendev.org/c/openstack/kolla/+/978812 | 13:37 |
| mnasiadka | bbezak frickler kevko mmalchuk gkoper jovial mattcrees dougszu darmach pabloclsn ravlew amir58118 r-krcek blanson[m] - meeting in 7 minutes | 13:53 |
| * fungi is also around for the additional agenda item | 13:56 | |
| tafkamax | I also added my little review item | 13:57 |
| thelounge87 | Umm how to add one here? | 13:58 |
| fungi | thelounge87: https://etherpad.opendev.org/p/KollaWhiteBoard | 13:59 |
| fungi | "additional agenda" section | 13:59 |
| mnasiadka | #startmeeting kolla | 14:00 |
| opendevmeet | Meeting started Wed Mar 4 14:00:01 2026 UTC and is due to finish in 60 minutes. The chair is mnasiadka. Information about MeetBot at http://wiki.debian.org/MeetBot. | 14:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 14:00 |
| opendevmeet | The meeting name has been set to 'kolla' | 14:00 |
| mnasiadka | #topic rollcall | 14:00 |
| mnasiadka | O/ | 14:00 |
| seunghunlee | o/ | 14:00 |
| bbezak | o/ | 14:00 |
| fungi | ahoy! | 14:00 |
| jovial | o/ | 14:00 |
| blanson[m] | hello ! | 14:01 |
| thelounge87 | Hi | 14:01 |
| thelounge87 | Added mine to ideas | 14:02 |
| mnasiadka | #topic agenda | 14:03 |
| mnasiadka | * CI status | 14:03 |
| mnasiadka | * Release tasks | 14:03 |
| mnasiadka | * Regular stable releases (first meeting in a month) | 14:03 |
| mnasiadka | * Current cycle planning | 14:03 |
| mnasiadka | * Additional agenda (from whiteboard) | 14:03 |
| mnasiadka | * Open discussion | 14:03 |
| mnasiadka | #topic CI status | 14:03 |
| mnasiadka | We’re getting there, stable branches older than 2025.1 need some love | 14:04 |
| mnasiadka | If there’s any volunteer to have a look in backports and fixing that - would be good | 14:04 |
| mnasiadka | Although fixing might require doing some work in requirements repo | 14:04 |
| mnasiadka | #topic Regular stable releases | 14:04 |
| mnasiadka | I think we’ll hold off until all stable branches are in healthy and green condition and post them | 14:05 |
| mnasiadka | #topic Current cycle planning | 14:05 |
| mnasiadka | So - I’ve been off for a month, so haven’t yet tried to find the features or other patches that are in reviewable and mergeable condition - if there are any - just put them on the whiteboard in the additional agenda section | 14:05 |
| mnasiadka | And we were mostly busy with getting the branches to a healthy state given all the testtools/setuptools madness | 14:06 |
| mnasiadka | Anybody wants to ask anything about their patches or things that need to be worked on this cycle? | 14:06 |
| thelounge87 | is there any idea to get back the backwards compatibility to haproxy and keepalived? | 14:07 |
| thelounge87 | as in any scope* | 14:07 |
| seunghunlee | I have two: https://review.opendev.org/c/openstack/kolla-ansible/+/956771 and https://review.opendev.org/c/openstack/kolla-ansible/+/953297 | 14:08 |
| mnasiadka | thelounge87: we’ve had enough of mariadb-clustercheck and have moved towards using proxysql as a specialised load balancer for MariaDB (especially for people needing multiple shards) - you have to live with it, ideally propose patches or at least bugs for functionality that you’re missing now | 14:08 |
| butjar | I wanted to ask if there is anything I can help with to advance VPNaaS (#link https://review.opendev.org/c/openstack/kolla-ansible/+/924575)? | 14:10 |
| mnasiadka | seunghunlee: I’ll have a look later this week in these patches, and I assume bbezak can help | 14:10 |
| seunghunlee | Thanks | 14:10 |
| bbezak | Sure | 14:10 |
| thelounge87 | So does that mean, its better to remove any existing deployment which doesn't use proxysql and redploy using proxysql? | 14:11 |
| mnasiadka | butjar: I think my privsep and privileged mode questions need to be answered | 14:11 |
| butjar | mnasiadka: I think this has been tackled by Vii | 14:11 |
| mnasiadka | thelounge87: that means an upgrade from mariadb-clustercheck to proxysql should be completely easy and we have done multiple of those in our downstream job - I’d like to understand your problems - but please create a bug in https://bugs.launchpad.net/kolla-ansible and let’s continue the discussion there | 14:12 |
| thelounge87 | okay | 14:12 |
| mnasiadka | butjar: comments are still unresolved | 14:13 |
| mnasiadka | That’s the minimum I need to take another look there | 14:13 |
| mnasiadka | But it might take a couple of weeks to get there | 14:13 |
| mnasiadka | Ok, let’s move on. | 14:13 |
| mnasiadka | #topic Additional agenda (from whiteboard) | 14:14 |
| mnasiadka | Let’s start with fungi’s items, he surely has some more interesting work to do :) | 14:14 |
| butjar | mnasiadka: Ok, THX I'll have a look at the comments | 14:14 |
| fungi | this is for "Bridging the Gap Flamingo Cycle Retrospective" | 14:14 |
| fungi | ildiko posted openstack-wide 2025.2 (flamingo) cycle retrospective contributor/maintainer survey results and metrics to openstack-discuss at the end of last year: | 14:14 |
| fungi | #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/XZZYFHMUCB3IZU5AXM366AY7WJXQMTBX/ Bridging the Gap Flamingo Cycle Retrospective Survey Results | 14:14 |
| fungi | #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/FD4JALJN7OB2YNBUVWCPZZ6YNJMIL2JT/ Bridging the Gap Flamingo Cycle Retrospective Metrics Analysis | 14:15 |
| fungi | i and the other community managers on the openinfra foundation staff have also been digging into team-specific details and i'm doing a round of outreach similar to last cycle, to go over how things may have changed | 14:15 |
| fungi | the kolla team had 1 contributor and 1 maintainer fill out the surveys, more would be preferable of course so i'll try to get them better publicized for the gazpacho round | 14:15 |
| fungi | the contributor was active for at least a year, also contributing to 2 or more other open source projects | 14:15 |
| fungi | ratings on the contributor survey questions were higher than the previous cycle, with timely reviews and actionable review feedback highest at 5 out of 5 while the rest were mostly ranked at 3 (useful test results came in at 4) | 14:15 |
| tafkamax | Regarding the fix for IDP https://review.opendev.org/c/openstack/kolla-ansible/+/975901 | 14:15 |
| tafkamax | It has grown rather large. Do we need to split this up in any way? Or what are the thoughts? It doesn't introduce new options IMO, so it should be backportable aswell. | 14:15 |
| fungi | the contributor additionally noted that irc discussions were hard to participate in due to concurrent bursts of review notifications in the channel | 14:15 |
| fungi | the new survey questions about priorities indicated the contributor followed development and prioritization through review-priority votes in gerrit and coverage in team meetings | 14:16 |
| fungi | by comparison, the maintainer who responded listed those same methods but also indicated they managed or communicated priorities through etherpad, gerrit change topics, custom review dashboard/query, and discussion in irc or mailing list | 14:16 |
| fungi | the rankings on the maintainer survey response were even higher, putting useful test results and comprehensiveness of contributor docs at 5 out of 5, with everything else a 4 | 14:16 |
| fungi | the maintainer indicated having no problems getting reviews on their own changes, but noted that many changes from contributors were lower priority, outside their expertise, not of interest, or they had insufficient time to review deeply | 14:16 |
| fungi | as for metrics, active reviewer count rose by 11% while active maintainer count fell by 31% in the flamingo cycle as compared to epoxy, but time to review/time to merge improved dramatically even with fewer maintainers | 14:16 |
| fungi | further, the count of new changes opened increased by 18% in flamingo, but maintainers continued to close 15% more changes than were opened, so the team seems to be more than keeping up with the overall review volume | 14:17 |
| fungi | the community managers have been distilling feedback from last year's discussions, and are working on compiling a concise set of techniques/recommendations for improving communication and efficiency, which i'll get into in a later meeting | 14:17 |
| fungi | we're also hoping to get some case studies done in concert with successful contributors and maintainers to highlight how specific practices and behaviors help them achieve better throughput, in order to try replicating those patterns | 14:17 |
| fungi | anyway, that was a quick dump, i know it's a lot to take in but i didn't want to eat up too much of your meeting, so i've put a followup on the agenda two weeks from now to give everyone time to digest and come up with questions or ideas | 14:17 |
| fungi | though i'm happy to answer any immediate feedback now if there's time | 14:17 |
| thelounge87 | so are there any plans to move away from IRC to some other channel of communication which is more popular? | 14:18 |
| fungi | is that a question for kolla leadership, or for the openstack technical committee? if the latter, you'd want to bring that up with them | 14:18 |
| mnasiadka | There’s aninitiative from OpenDev team to make Matrix feasible for move, but I don’t think it’s going to happen extremely soon | 14:18 |
| mnasiadka | Yeah, it would be best to bring it up with TC | 14:19 |
| thelounge87 | I personally believe at least adding a guide on how to install an irc client would be really helpful to the contributors guide. | 14:19 |
| fungi | yeah, i mean, some parts of openstack are already using the opendev.org matrix homeserver, like the ops community | 14:19 |
| tafkamax | I am using the matrix over the irc bridge. | 14:20 |
| thelounge87 | I am currently using thelounge snap. | 14:20 |
| fungi | i'm using weechat connected to 4 irc networks, matrix, and 3 slack accounts | 14:20 |
| mnasiadka | fungi: I’d like to tackle the notification bursts - on IRC that’s probably as easy as moving them to a separate channel… In future on Matrix - is there a way client can filter out bot messages? | 14:20 |
| fungi | it would basically be the same approach on irc or matrix: set up a separate channel/room, or otherwise reduce the volume of reporting | 14:21 |
| fungi | i've seen both approaches used, just depends on the team's overall preferences | 14:21 |
| thelounge87 | and a way to guess what actually is a bot message :) | 14:21 |
| mnasiadka | Ok, we’ll discuss which approach is better | 14:22 |
| fungi | well, on matrix the gerritbot does use the notification type rather than normal user comments | 14:22 |
| fungi | one thing we could look into for the irc gerritbot is using irc notices instead of comments | 14:23 |
| fungi | s/comments/messages/ | 14:23 |
| tafkamax | I have to say. I don't mind the notifications, but on the OpenDev matrix channel the notification type events are smaller and more comprehendible. So easier on the eyes. | 14:23 |
| fungi | right now, our statusbot notices use irc notice type instead of normal messages | 14:24 |
| fungi | so it should be doable | 14:24 |
| fungi | though i don't know if it's any less verbose | 14:24 |
| mnasiadka | fungi: thanks for your work with bridging the gap initiative, the feedback is really useful and I guess mainly positive | 14:24 |
| fungi | yw, we can cover this more in a couple of weeks (i have a conflicting conference call during the next kolla meeting) | 14:25 |
| mnasiadka | fungi: I think on IRC the obvious solution is make the notifications less verbose or move them to a different channel - we’ll think about it | 14:25 |
| thelounge87 | My feature/suggestion/request/code: | 14:26 |
| thelounge87 | ``` | 14:26 |
| thelounge87 | Multiregion Support. With a seperate identity database (Keystone Database) that asynchronously syncs over VPN with multiple databases. Using the same base Mariadb Image. But a different container (Calling it Mariadb-identity). Instead of the current recommended architecture which is essentially a single point of failure as all the db are | 14:26 |
| thelounge87 | concentrated in a single region and connected through VPN? | 14:26 |
| thelounge87 | ``` | 14:26 |
| thelounge87 | This is my feature request and made a commit regarding the same. But will modify it with proxysql. Thoughts on this? | 14:26 |
| tafkamax | I don't know how it looks like in the other clients, but in matrix, the irc review URL is fully printed, while in native matrix it is a LinkText, with the title being clickable. | 14:26 |
| tafkamax | (Example being the OpenDev matrix channel again) | 14:27 |
| mnasiadka | thelounge87: I think it would be useful to discuss that, but given I’m the only core around on this meeting - a better venue would be Project Teams Gathering in April | 14:28 |
| pabloclsn_ | hi everyone, how can we make this : https://review.opendev.org/c/openstack/kolla/+/924302 moove ? why is it stuck ? also in openstack 2025.2 with kolla-ansible in 2025.2 we cannot make maria_backup anymore i created two fixes : https://review.opendev.org/c/openstack/kolla/+/978300 https://review.opendev.org/c/openstack/kolla-ansible/+/978299 | 14:29 |
| mnasiadka | pabloclsn_: there was a question about the vpn agent earlier, sync with butjar | 14:30 |
| pabloclsn_ | oh didn't see it thanks :) | 14:30 |
| thelounge87 | So gotta wait till april ? or should I make an email thread? | 14:30 |
| mnasiadka | Around the two others - the Kolla one is failing CI, although I added them both to my list and will try for at least some initial comments | 14:31 |
| mnasiadka | thelounge87: We would prefer for you to wait for April, OpenStack is mainly close to a release - Kolla is a cycle-trailing project and we have some more time for a release (a couple weeks after the coordinated release) - but it won’t be the time for merging new features | 14:32 |
| thelounge87 | okay | 14:32 |
| blanson[m] | mnasiadka: I was gonna ask, we have a feature freeze window right ? I think I saw something about that somewhere ? so I don't send patches right now if they're going to be new feature (for next cycle) | 14:33 |
| mnasiadka | blanson[m]: basically the rule up to now was that if you can find two cores who agree to review your patch to completion - that’s your feature freeze exception | 14:34 |
| mnasiadka | But with only three cores active, and me and Bartosz being a bit tied to our downstream work for at least the next 3-4 weeks - it might be complicated | 14:34 |
| KurtB | I added an Idea in the whiteboard: Provide selinux support.There's a need for this and I'd like to do this work. Feedback and opinions welcome. | 14:34 |
| thelounge87 | 2 of 3 core maintainers. Thats a tight knit community | 14:36 |
| blanson[m] | mnasiadka: thanks ! yh I figured you guys were busy enough as it was. I'm gonna hold some of them for next cycle :) | 14:36 |
| butjar | pabloclsn_: I sent u a DM | 14:36 |
| opendevreview | Seunghun Lee proposed openstack/kolla-ansible master: Improve ProxySQL test https://review.opendev.org/c/openstack/kolla-ansible/+/978221 | 14:37 |
| mnasiadka | Ok then, the rest of additional agenda are Gerrit links so we’ll have a look there together with bbezak when time permits | 14:37 |
| mnasiadka | #topic Open discussion | 14:37 |
| mnasiadka | Anybody anything? | 14:37 |
| thelounge87 | 1. I want more documentation on proxysql | 14:38 |
| mnasiadka | blanson[m]: if it’s very important for you we can probably make it work, but if you can wait a bit - then it might be better | 14:38 |
| mnasiadka | thelounge87: Are you willing to write it? | 14:38 |
| thelounge87 | If I understood the thought behind it, sure | 14:38 |
| thelounge87 | but we have seunghunlee the core maintainer of that right? | 14:39 |
| blanson[m] | no no, it was about the keystone encryption stuff, but it requires some work + it's quite heavy to review imo | 14:39 |
| blanson[m] | so it's probbly better to hold off until the next cycle, at least the review part of it | 14:39 |
| mnasiadka | thelounge87: seunghunlee is a community member and my co-worker, don’t expect people to help you for free just because this is an open source project | 14:40 |
| KurtB | Jumped the gun on the SELinux note above. We can let that cook for a while and give people time to think about it while I work up an example for review. | 14:40 |
| seunghunlee | Not sure when I became a core of that. The patch I'm working on https://review.opendev.org/c/openstack/kolla-ansible/+/974712 might bring some docs. idk | 14:40 |
| thelounge87 | Okay then, I'll write out my understanding of proxysql. And You can check the PR of it. | 14:40 |
| seunghunlee | thelounge87: This is open source project and we welcome your contributions! | 14:40 |
| thelounge87 | https://review.opendev.org/c/openstack/kolla-ansible/+/962479 I was going off by who made this commit so yeah. | 14:41 |
| mnasiadka | KurtB: we can discuss that in two weeks, there should be more people available for some more insights - and you probably will have more kno | 14:41 |
| mnasiadka | More ideas | 14:41 |
| KurtB | True. Thanks. | 14:41 |
| mnasiadka | Ok then, since all is clear - I’m closing the meeting. | 14:42 |
| mnasiadka | Thank you all for coming! | 14:42 |
| mnasiadka | #endmeeting | 14:42 |
| opendevmeet | Meeting ended Wed Mar 4 14:42:30 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 14:42 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/kolla/2026/kolla.2026-03-04-14.00.html | 14:42 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/kolla/2026/kolla.2026-03-04-14.00.txt | 14:42 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/kolla/2026/kolla.2026-03-04-14.00.log.html | 14:42 |
| blanson[m] | thanks ! few that was a dense meeting | 14:43 |
| tafkamax | thanks | 14:43 |
| blanson[m] | s/few/phewww/ | 14:43 |
| tafkamax | let's pray that the CI is gentle on us now | 14:43 |
| tafkamax | I saw a lot of verification fails after the merge. | 14:44 |
| tafkamax | In opendev matrix channel they talked about post merge verfication changes. | 14:44 |
| tafkamax | That was written yesterday | 14:44 |
| tafkamax | > if anyone else finds users confused about unexplained merge failures it looks like github was having an outage today and that impacted some merge requests for some jobs. | 14:44 |
| fungi | the ci gods demand frequent sacrifices | 14:49 |
| fungi | and yeah, we saw a bunch of merge failures due to most devstack jobs consuming novnc from git source, so when zuul was unable to check github to determine the current branch state there it couldn't proceed | 14:50 |
| opendevreview | Antony Messerli proposed openstack/kolla-ansible master: ovn-db: add support for ovn-northd extra command-line arguments https://review.opendev.org/c/openstack/kolla-ansible/+/977469 | 15:34 |
| opendevreview | Seunghun Lee proposed openstack/kolla-ansible master: Improve ProxySQL test https://review.opendev.org/c/openstack/kolla-ansible/+/978221 | 16:53 |
| opendevreview | Seunghun Lee proposed openstack/kolla master: WIP: Update RPM repos to use RL10 version if available https://review.opendev.org/c/openstack/kolla/+/978911 | 16:57 |
| opendevreview | Seunghun Lee proposed openstack/kolla-ansible master: Improve ProxySQL test https://review.opendev.org/c/openstack/kolla-ansible/+/978221 | 17:02 |
| -opendevstatus- NOTICE: The Gerrit service on review.opendev.org will be offline momentarily for a software upgrade | 18:11 | |
| *** r3ap3r_ is now known as r3ap3r | 18:16 | |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!