-@gerrit:opendev.org- Ian Wienand proposed: | 00:16 | |
- [zuul/zuul-jobs] 866953: ensure-kubernetes: add microk8s support https://review.opendev.org/c/zuul/zuul-jobs/+/866953 | ||
- [zuul/zuul-jobs] 867063: [wip] zuul-jobs-test-registry-buildset-registry-k8s-microk8s https://review.opendev.org/c/zuul/zuul-jobs/+/867063 | ||
-@gerrit:opendev.org- Ian Wienand proposed: | 00:24 | |
- [zuul/zuul-jobs] 866953: ensure-kubernetes: add microk8s support https://review.opendev.org/c/zuul/zuul-jobs/+/866953 | ||
- [zuul/zuul-jobs] 867063: [wip] zuul-jobs-test-registry-buildset-registry-k8s-microk8s https://review.opendev.org/c/zuul/zuul-jobs/+/867063 | ||
-@gerrit:opendev.org- Ian Wienand proposed: | 00:35 | |
- [zuul/zuul-jobs] 866953: ensure-kubernetes: add microk8s support https://review.opendev.org/c/zuul/zuul-jobs/+/866953 | ||
- [zuul/zuul-jobs] 867063: [wip] zuul-jobs-test-registry-buildset-registry-k8s-microk8s https://review.opendev.org/c/zuul/zuul-jobs/+/867063 | ||
-@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul-jobs] 867063: use-buildset-registry: support microk8s https://review.opendev.org/c/zuul/zuul-jobs/+/867063 | 01:02 | |
-@gerrit:opendev.org- Tim Beermann proposed: [zuul/zuul-jobs] 866679: Add yamllint job. https://review.opendev.org/c/zuul/zuul-jobs/+/866679 | 06:37 | |
-@gerrit:opendev.org- Tim Beermann proposed: [zuul/zuul-jobs] 866679: Add yamllint job. https://review.opendev.org/c/zuul/zuul-jobs/+/866679 | 06:47 | |
-@gerrit:opendev.org- Tim Beermann proposed: [zuul/zuul-jobs] 866679: Add yamllint job. https://review.opendev.org/c/zuul/zuul-jobs/+/866679 | 06:59 | |
-@gerrit:opendev.org- Felix Edel proposed: [zuul/zuul] 866863: Store pause and resume events on the build and report them https://review.opendev.org/c/zuul/zuul/+/866863 | 08:02 | |
-@gerrit:opendev.org- Felix Edel proposed: [zuul/zuul] 866863: Store pause and resume events on the build and report them https://review.opendev.org/c/zuul/zuul/+/866863 | 08:08 | |
-@gerrit:opendev.org- Felix Edel proposed: [zuul/zuul] 866863: Store pause and resume events on the build and report them https://review.opendev.org/c/zuul/zuul/+/866863 | 08:11 | |
-@gerrit:opendev.org- Zuul merged on behalf of Simon Westphahl: [zuul/zuul] 865687: Consider queue settings for topic dependencies https://review.opendev.org/c/zuul/zuul/+/865687 | 15:07 | |
@clarkb:matrix.org | Hello Zuulians. I have been asked to remind you all that the OpenInfra Summit CFP closes in four weeks. The summit runs June 13-15 in Vancouver, BC Canada. If you would like to talk about your use of Zuul or other relevant technologies feel free to submit a proposal here: https://cfp.openinfra.dev/app/vancouver-2023/19. You can learn more about the event here: https://openinfra.dev/summit/vancouver-2023 | 16:42 |
---|---|---|
-@gerrit:opendev.org- Artem Goncharov proposed: [zuul/zuul] 859939: Implement Gitea driver https://review.opendev.org/c/zuul/zuul/+/859939 | 17:11 | |
-@gerrit:opendev.org- Artem Goncharov proposed: [zuul/zuul] 859940: Implement tests of the gitea driver https://review.opendev.org/c/zuul/zuul/+/859940 | 17:11 | |
-@gerrit:opendev.org- Ade Lee proposed: [zuul/zuul-jobs] 866881: Add ubuntu to enable-fips role https://review.opendev.org/c/zuul/zuul-jobs/+/866881 | 17:12 | |
-@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] 865923: Avoid acquiring pipeline locks in manager postConfig https://review.opendev.org/c/zuul/zuul/+/865923 | 18:07 | |
-@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 867175: Reuse queue items after reconfiguration https://review.opendev.org/c/zuul/zuul/+/867175 | 21:20 | |
@iwienand:matrix.org | ``` | 23:10 |
2022-12-13 23:09:51,449 ERROR nodepool.builder.CleanupWorker.0: Unable to delete image ubuntu-xenial-1670176248 from rax-iad: | ||
Traceback (most recent call last): | ||
File "/usr/local/lib/python3.11/site-packages/nodepool/builder.py", line 326, in _deleteUpload | ||
manager.deleteImage(upload.external_name, | ||
File "/usr/local/lib/python3.11/site-packages/nodepool/driver/openstack/provider.py", line 223, in deleteImage | ||
return self._client.delete_image(dict(id=id)) | ||
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | ||
File "/usr/local/lib/python3.11/site-packages/openstack/cloud/_image.py", line 223, in delete_image | ||
self.image._IMAGE_OBJECT_KEY in image.properties | ||
^^^^^^^^^^^^^^^^ | ||
@iwienand:matrix.org | this is on nb02, and looks like something openstackclient-y | 23:11 |
@iwienand:matrix.org | it's in a pretty tight loop on that | 23:11 |
@clarkb:matrix.org | hrm that could be related to the openstacksdk update. | 23:12 |
@clarkb:matrix.org | When I spot checked it it looked like images were deleting and uploading properly though | 23:12 |
@clarkb:matrix.org | ianw: is that traceback complete? | 23:13 |
@iwienand:matrix.org | that traceback is all we're getting, yep | 23:14 |
@clarkb:matrix.org | `AttributeError: 'dict' object has no attribute 'properties'` is the bit that is missing | 23:15 |
@iwienand:matrix.org | ohhh, sorry, copy-paste fail | 23:15 |
@clarkb:matrix.org | and an image list shows that we are leaking them in the cloud. | 23:16 |
@iwienand:matrix.org | https://review.opendev.org/c/openstack/openstacksdk/+/823092 seems to be where it changed to looking at .properties | 23:17 |
@clarkb:matrix.org | looking at that code this just seems to be broken | 23:17 |
@clarkb:matrix.org | get_image is returning a dict and this is exploding because dicts don't have a .properties | 23:18 |
@iwienand:matrix.org | well it might be something like the other bug with rax where ".properties" isn't set due to -EVERYOLD or something. it only seems to be rax | 23:18 |
@clarkb:matrix.org | yes, though the error says image is a dict and dicts don't have a .properties | 23:19 |
@clarkb:matrix.org | if they are modifying dict objects with a .properties... | 23:19 |
@jim:acmegating.com | remember that shade used to have munch objects which were dicts that behaved like objects with accessors for fields | 23:20 |
@jim:acmegating.com | not sure what the current status is on that with the latest sdk update | 23:20 |
@clarkb:matrix.org | we should be able to run a little script that does a cloud.get_image() and then see what the results are | 23:21 |
@jim:acmegating.com | i'll do that. i should be able to re-purpose my test environment | 23:22 |
@iwienand:matrix.org | i also have some test bits left over from the last openstacksdk rax breakage :) | 23:23 |
@clarkb:matrix.org | ```import openstack | 23:24 |
@clarkb:matrix.org | * ```import openstack | 23:24 |
``` | ||
@clarkb:matrix.org | I don't know how to do rich text | 23:24 |
@jim:acmegating.com | Clark: to block quote something, 3 backticks on one line to start, and the same to end | 23:25 |
@clarkb:matrix.org | * ```import openstack | 23:25 |
conn = openstack.connect(cloud='rax') | ||
``` | ||
@clarkb:matrix.org | * ``` | 23:26 |
import openstack | ||
#TODO how to set region? | ||
conn = openstack.connect(cloud='rax') | ||
image = conn.get_image('imageuuidorname') | ||
print(image) | ||
``` | ||
@clarkb:matrix.org | I think this image management path is fairly unique to rax too beacuse they do the swift backed images whereas everyone else is direct upload to glance | 23:28 |
@iwienand:matrix.org | ``` | 23:29 |
r = cloud.image.get_image('6604871b-3d19-4727-8b05-2501a4a8afad') | ||
print(r.properties) | ||
``` | ||
@iwienand:matrix.org | works | 23:29 |
@iwienand:matrix.org | have to quickly pick up kid from shop, back in 5 | 23:29 |
@jim:acmegating.com | that's not the equivalent call; it's cloud.image.get_image({'id': '660...'}) | 23:35 |
@jim:acmegating.com | get_image({'id': '6604871b-3d19-4727-8b05-2501a4a8afad'}) | 23:35 |
@jim:acmegating.com | that reproduces the error | 23:35 |
@clarkb:matrix.org | look at the definition of _get_entity. I think it short circuits and returns what was passed if you pass it a dict like that | 23:36 |
@clarkb:matrix.org | ianw's case goes throughactual lookup | 23:37 |
@clarkb:matrix.org | so maybe we need to call it like ianw in nodepool? | 23:37 |
@jim:acmegating.com | yes, this looks like it may be an api change from sdk | 23:37 |
@clarkb:matrix.org | yeah. But the fix should be easy I guess? | 23:38 |
@clarkb:matrix.org | we could even hand patch nodepool on nb01/nb02 and restart the containers to see if it works | 23:38 |
@clarkb:matrix.org | the new thing should at least be backward compatible though I think | 23:39 |
@clarkb:matrix.org | but we may want to set openstacksdk>=0.103.0 anyway | 23:39 |
@iwienand:matrix.org | get_image just says | 23:39 |
@iwienand:matrix.org | image – The value can be the ID of a image or a Image instance. | 23:39 |
@iwienand:matrix.org | i wonder if it ever said a dict | 23:40 |
@jim:acmegating.com | Clark: agree the fix is simple; unsure about backwards compat, the version min may be safest | 23:40 |
@iwienand:matrix.org | let me hand-patch nb02 and see | 23:41 |
@clarkb:matrix.org | ianw you will need to docker-compose stop then start it after you edit it | 23:42 |
@iwienand:matrix.org | i just hup'd it | 23:43 |
@jim:acmegating.com | i'm not sure it's worth digging in deeper, but i think that these methods probably haven't really changed much -- except that previously they were simple enough that they would work with {'id': '...'} as input, but now they expect a more fully-formed object as input | 23:43 |
@iwienand:matrix.org | it's deleting a bunch now with no warnings | 23:43 |
@jim:acmegating.com | * i'm not sure it's worth digging in deeper, but i think that these methods probably haven't really changed much -- except that previously they were simple enough that they would work with {'id': '...'} as input, but now they expect a more fully-formed object as input (by accessing the '.properties' attr) | 23:43 |
@iwienand:matrix.org | https://paste.opendev.org/show/bXkJefwBIGI94wgIdEDq/ | 23:43 |
@jim:acmegating.com | so basically, there previously was an old informal api where you could pass in a "good enough" object, and now that isn't supported anymore so we should fix nodepool. | 23:44 |
@clarkb:matrix.org | corvus: there is a change from may that converts a number of calls that similarly called get_image to just pass in the id bare | 23:44 |
@clarkb:matrix.org | in openstacksdk I mean | 23:44 |
@clarkb:matrix.org | ianw: did you want to propose the fix since you wrote the change on nb02? | 23:44 |
@clarkb:matrix.org | ianw: if so can you update the requirements to require openstacksdk>=0.103.0? | 23:44 |
@jim:acmegating.com | Clark: ianw may i propose the fix? | 23:44 |
@clarkb:matrix.org | corvus: oh sure. I'm happy for anyone to do it :) | 23:45 |
@iwienand:matrix.org | sure | 23:45 |
@jim:acmegating.com | thx, i want to stack the adapter on it at the same time | 23:45 |
@iwienand:matrix.org | ok, well happily it was just the deletion path, i was a bit worried at first with all the flooding messages | 23:45 |
@iwienand:matrix.org | now back to why tox-py38 is failing on dib... | 23:46 |
@clarkb:matrix.org | spot checking the cleanup it is doing now I see images it says are deleting are no longer present and they were in my early image-list'ing | 23:47 |
@clarkb:matrix.org | But there are still a number of deleting images present so this may take some time I guess | 23:47 |
@clarkb:matrix.org | doing a somewhat naive grep for similar patterns in the openstack driver produces no other results which hopefully means this is the only place we're broken like this | 23:48 |
@clarkb:matrix.org | I have now confirmed via `wc -l` that the numer of images is steadily dropping so ya just needs time | 23:49 |
-@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 23:50 | |
- [zuul/nodepool] 862522: Convert openstack driver to statemachine https://review.opendev.org/c/zuul/nodepool/+/862522 | ||
- [zuul/nodepool] 867578: Fix openstack image deletion with newer sdk https://review.opendev.org/c/zuul/nodepool/+/867578 | ||
@jim:acmegating.com | ianw: Clark ^ | 23:51 |
@jim:acmegating.com | Clark: would you mind re-adding your +2 on the statemachine change while it's fresh in your mind? (should be the only diff) | 23:51 |
@clarkb:matrix.org | corvus: ya | 23:51 |
@iwienand:matrix.org | ... i should spend some time with that too, sorry it's mostly passed me by | 23:53 |
@clarkb:matrix.org | gerrit did confirm that was the only diff | 23:53 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!