| opendevreview | OpenStack Proposal Bot proposed openstack/openstack-manuals master: Imported Translations from Zanata https://review.opendev.org/c/openstack/openstack-manuals/+/947180 | 03:31 |
|---|---|---|
| *** dalees8 is now known as dalees | 06:20 | |
| jwysogla | fungi Hello o/ we recently started a new repository in telemetry called Aetos https://opendev.org/openstack/aetos . There was a concern last Friday, if we're allowed to use that name. We had a brief conversation with gouthamr previously and I was pointed to you also. I checked pypi, debian and fedora packages and there is no collision. Are we good or is there something else that should be checked? | 07:41 |
| fungi | jwysogla: i see a few open source projects just doing web searches, but none seem significant or active; probably the only other concern would be if there are commercial software products whose owners might try to claim trademark infringement, though i'm not spotting any obvious ones | 13:15 |
| fungi | what was the specific concern raised with being able to use the name? | 13:15 |
| jwysogla | The exact message from Takashi was: <tkajinam> mrunge_, mmagr__ I wonder if anyone talked with foundation staffs about the naming "aetos" ? AFAIR the foundation does some trademark checks to determine if the project name is acceptable (and in the past some projects needed renaming due to that) maybe that was discussed with TC when the new project was proposed in governance channel. | 13:18 |
| jwysogla | So I just want to know if there is something else we need to check around the naming. I agree, that there doesn't seem anything obvious with the name "Aetos" from the googling I've done. | 13:20 |
| fungi | we've done trademark checking on project names, but generally not at the level of repository names. since this is just going to be part of the telemetry project we wouldn't be marketing/promoting it with the name aetos anyway | 13:23 |
| fungi | from a foundation-level marketing perspective, it's just part of "openstack telemetry" | 13:23 |
| jwysogla | Alright. Thank you | 13:24 |
| tkajinam | makes clear sense | 13:24 |
| fungi | (we did perform trademark searches on "ceilometer" once upon a time, but that's because the whole team/project used to go by that name before it was changed to "telemetry") | 13:24 |
| fungi | for a more pointed example, i'm pretty sure we didn't perform any trademark searches for gnocchi back when it got added to the telemetry project (before it moved out of openstack) | 13:27 |
| cardoe | Anyone know if sshnaidm is usually on IRC? | 14:41 |
| fungi | haven't seen them in a while, not since tripleo dissolved | 15:17 |
| tkajinam | cardoe, afaik he is still in Red Hat but he moved his focus from OpenStack to OpenShift some time ago. you can drop an email or ask someone from Red Hat to connect you with him | 15:22 |
| tkajinam | if needed | 15:22 |
| cardoe | Just trying to get some changes in on ansible-collection-openstack | 15:23 |
| cardoe | and it seems that him and gtema are needed to review and +W | 15:23 |
| tkajinam | https://review.opendev.org/admin/groups/0e01228e912733e8b9a8d957631e41665aa0ffbd,members | 15:25 |
| tkajinam | I'm afraid gtema is the only active person from the core group nowadays | 15:25 |
| frickler | cardoe: which changes? I think any sdk core should be able to approve things there, including fellow infra-rooters and myself | 15:27 |
| cardoe | Well not looking to push things through specially. But how do we get to a sustainable place? | 15:28 |
| cardoe | The tests have been broken since February so I'm trying to fix those. And then I was going to try and land some folks changes and push a release. | 15:30 |
| gtema | Will do so tonight | 16:13 |
| noonedeadpunk | there was another folks asking for reviews today btw... | 16:16 |
| noonedeadpunk | aha, cardoe also on that :) | 16:17 |
| noonedeadpunk | But I've added a topic on their behalf to the next TC meeting | 16:18 |
| noonedeadpunk | they are planning to attend it as well | 16:18 |
| fungi | yeah, when looking at core group membership on gerrit, don't forget to check a group's list of included groups too | 16:21 |
| frickler | anyway fixing the CI would seem to need to be done first anyway. looks like sagi was active on https://review.opendev.org/c/openstack/ansible-collections-openstack/+/942397 recently, but that patch doesn't seem to be enough (anymore). it would also be interesting to find out whether these are really just bad tests or whether there might be recent regressions in openstack that have otherwise went | 16:32 |
| frickler | unnoticed so far | 16:32 |
| frickler | cardoe: simondodsley: I guess the question to you two would also be whether you "only" want to get your features merged, or whether you'd also be willing to commit to long(er) term maintenance of the project | 16:39 |
| sean-k-mooney | i know absolutely nothign about that repo btu the test is failing to create a recordset in designate and designate sames to be returnign a 400? | 16:41 |
| sean-k-mooney | GET /v2/zones/9da9e374-5870-4c88-a66c-0b98266bc063/recordsets/testrecordset.test.dns.zone. => generated 165 bytes in 7 msecs (HTTP/1.1 400) 4 headers in 166 bytes (1 switches on core 0) | 16:41 |
| sean-k-mooney | im not exactly sure sematically how a http get is a client errror let alone a 400 | 16:41 |
| sean-k-mooney | but there is nithign obvious in the desginate logs but maybe reach out to them and them might be abel to help | 16:42 |
| simondodsley | frickler: I'm potentially open to doing longer term support for this, but I don't have a lot of spare time to allocate at the moment | 16:43 |
| johnsom | Do you have a link to the failing job? | 16:43 |
| sean-k-mooney | i think its tecnially a diffent failing one but https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_a7e/openstack/a7e8667df9bd4ba8851a9f35c8a5f951/controller/logs/screen-designate-api.txt | 16:43 |
| sean-k-mooney | thats failing on https://review.opendev.org/c/openstack/ansible-collections-openstack/+/947597 | 16:43 |
| simondodsley | this is an example: https://zuul.opendev.org/t/openstack/build/46367d6da6f34363a552a7d57b3e0eda | 16:43 |
| johnsom | It looks like the test is querying for a zone that is not yet created. The central log shows it being created slightly after the API HTTP 400 log. | 16:49 |
| clarkb | the request to create the zone is probably async and requestors need to wait for it to be available? | 16:50 |
| johnsom | Yes | 16:50 |
| clarkb | can probably just do a retry as long as the response is a 400 | 16:50 |
| clarkb | and give up after a minute or some reasonable timeout | 16:50 |
| johnsom | Here https://review.opendev.org/c/openstack/ansible-collections-openstack/+/947597/1/ci/roles/recordset/tasks/main.yml it is not waiting for the zone to be created before it goes on to attempting to create a recordset. There is propagation time (i.e. for BIND to create the zone, etc.) before the zone will go ACTIVE. | 16:55 |
| sean-k-mooney | i wonder if the timing changed with the mvoe to noble | 16:59 |
| johnsom | Well, the nodepool hosts have been slower recently as well, likely due to the end of cycle push. | 17:00 |
| cardoe | So I've been adding some "untils" | 17:04 |
| johnsom | cardoe Yeah, so I think we need one on the zone create above as well. | 17:05 |
| cardoe | Well "Ensure DNS zone" from the logs has status ACTIVE | 17:07 |
| cardoe | Looking at https://zuul.opendev.org/t/openstack/build/a7e8667df9bd4ba8851a9f35c8a5f951 | 17:07 |
| johnsom | The create came in at 19:14:37.730948 and went active 19:14:44.058569. The GET for it was at 19:14:37.120977 | 17:07 |
| cardoe | yeah I'm just not sure what to trigger on because the task sees it created | 17:09 |
| cardoe | it possible that's the GET internally in the module? | 17:09 |
| johnsom | The failed GET call was for ansible.test.zone. which is different from the test.dns.zone. I see in that log | 17:10 |
| johnsom | Though it's a bit hard to see in that log. It's reused in a few places it seems. | 17:11 |
| cardoe | So my change fails now because of my until, because when it actually succeeds there is no change and the next test below is checking for if it changed. | 17:12 |
| cardoe | The only the first call to create it sets changed=True, otherwise it's already "created" therefore changed=False | 17:12 |
| johnsom | Yeah, I can see that | 17:12 |
| cardoe | But the rest of the check fails because it wants changed=True and then it wants to read data about it but that data is only populated when its ACTIVE and not PENDING | 17:13 |
| cardoe | So I was going to drop the "is changed" check and instead add another task to ensure that recordset doesn't exist? | 17:14 |
| opendevreview | Vladimir Kozhukalov proposed openstack/governance master: Retire openstack-helm-infra repo https://review.opendev.org/c/openstack/governance/+/947611 | 17:15 |
| cardoe | frickler: I mean I'll probably help care and feed it going forward for a while. | 17:59 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/governance master: Deprecate openstack-ansible-tests repo https://review.opendev.org/c/openstack/governance/+/947622 | 19:05 |
| cardoe | Yeah sure enough something is broken in Designate. | 20:04 |
| johnsom | cardoe Is it Designate or the Designate code in this ansible collections? If you think it is Designate, I am happy to dig through the Designate logs. | 20:14 |
| cardoe | Well I have no idea. It attempts to create a recordset with some records. The recordset goes active (my change made it wait for that) but the list of records that come back is empty while its expecting some in there now. | 20:15 |
| cardoe | I've looked through some of the logs and designate-central shows the recordset being made. | 20:17 |
| cardoe | the ansible module uses openstacksdk so I dunno if something changed there | 20:17 |
| cardoe | https://zuul.opendev.org/t/openstack/build/acdfc858d3d64abb99b0da0e50151383/log/tox/ansible_latest/1-commands[0].log#41411 so what's coming back is empty. | 20:19 |
| cardoe | I'm assuming that designate is accepting that since the recordset is being created and going ACTIVE | 20:20 |
| cardoe | https://zuul.opendev.org/t/openstack/build/acdfc858d3d64abb99b0da0e50151383/log/controller/logs/screen-designate-api.txt#163 that's where the API looks like it accepted that. | 20:22 |
| johnsom | This is an interesting one. I | 21:03 |
| johnsom | I am not sure what is going on. There isn't enough detail in those logs. I am going to have to setup a local devstack and try some things. | 21:05 |
| johnsom | Ok, so something odd is happening in the SDK or ansible. I ran this scenario on devstack via CLI and even the "PENDING" GET calls have the records populated, but in your run they are blank there too. | 22:00 |
| johnsom | Hmm, SDK seems to be working fine too | 22:21 |
| cardoe | So something in the module is reading the wrong fields from the SDK. Mind pasting me the SDK usage? Or what you see? | 22:31 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!