*** jawad_axd has joined #openstack-glance | 00:15 | |
*** jawad_axd has quit IRC | 00:50 | |
*** jawad_axd has joined #openstack-glance | 02:06 | |
*** jawad_axd has quit IRC | 02:41 | |
*** jawad_axd has joined #openstack-glance | 03:58 | |
*** jawad_axd has quit IRC | 04:31 | |
*** evrardjp has quit IRC | 04:33 | |
*** evrardjp has joined #openstack-glance | 04:33 | |
*** ratailor has joined #openstack-glance | 05:12 | |
*** udesale has joined #openstack-glance | 05:40 | |
*** jawad_axd has joined #openstack-glance | 06:09 | |
*** jawad_axd has quit IRC | 06:14 | |
*** jawad_axd has joined #openstack-glance | 07:00 | |
*** ratailor has quit IRC | 07:19 | |
*** ratailor has joined #openstack-glance | 07:19 | |
*** jawad_axd has quit IRC | 07:22 | |
*** jawad_axd has joined #openstack-glance | 07:22 | |
*** bhagyashris|afk is now known as bhagyashris | 07:27 | |
*** tosky has joined #openstack-glance | 07:35 | |
*** udesale has quit IRC | 09:07 | |
*** udesale has joined #openstack-glance | 09:15 | |
*** priteau has joined #openstack-glance | 10:03 | |
openstackgerrit | Abhishek Kekane proposed openstack/glance master: Add image_set_property_atomic() helper https://review.opendev.org/737868 | 10:33 |
---|---|---|
openstackgerrit | Abhishek Kekane proposed openstack/glance master: WIP: Fix race condition in copy image operation https://review.opendev.org/737596 | 10:33 |
* abhishekk signing out for the day | 10:38 | |
*** udesale has quit IRC | 12:29 | |
*** mordred has quit IRC | 13:34 | |
*** udesale has joined #openstack-glance | 13:47 | |
openstackgerrit | Dan Smith proposed openstack/glance master: Add image_set_property_atomic() helper https://review.opendev.org/737868 | 14:11 |
openstackgerrit | Dan Smith proposed openstack/glance master: WIP: Fix race condition in copy image operation https://review.opendev.org/737596 | 14:11 |
*** ratailor has quit IRC | 14:36 | |
*** priteau has quit IRC | 14:44 | |
*** priteau has joined #openstack-glance | 14:47 | |
*** priteau has quit IRC | 14:52 | |
*** priteau has joined #openstack-glance | 15:26 | |
*** priteau has quit IRC | 15:30 | |
*** priteau has joined #openstack-glance | 15:30 | |
*** mordred has joined #openstack-glance | 15:36 | |
*** jawad_axd has quit IRC | 15:36 | |
openstackgerrit | Merged openstack/glance_store master: Don't allow image creation with encrypted nfs volumes https://review.opendev.org/732506 | 15:51 |
*** k_mouza has joined #openstack-glance | 15:58 | |
k_mouza | good afternoon all! I was wondering if there's a way to move existing glance images from a file store to a RBD one on ceph? Is there a way to migrate them over without downloading and re-importing ? | 16:00 |
*** k_mouza has quit IRC | 16:14 | |
dansmith | jokke: around by chance/ | 16:18 |
*** k_mouza has joined #openstack-glance | 16:19 | |
dansmith | k_mouza: https://docs.openstack.org/glance/latest/admin/interoperable-image-import.html#configuring-the-copy-image-method | 16:20 |
*** k_mouza has quit IRC | 16:22 | |
*** k_mouza has joined #openstack-glance | 16:22 | |
mordred | dansmith: I updated the sdk patch to hopefully not add OpenStack-image-import-methods to properties anymore | 16:26 |
dansmith | mordred: cool, thanks | 16:26 |
mordred | dansmith: I don't yet know *why* it went there - that's explicity unexpected, but for now we can decouple this from solving that general problem | 16:26 |
dansmith | okay, I assume you just pushed that up recently? | 16:27 |
*** mordred has quit IRC | 16:33 | |
*** gyee has joined #openstack-glance | 16:35 | |
*** mordred has joined #openstack-glance | 16:38 | |
*** k_mouza has quit IRC | 16:39 | |
openstackgerrit | Dan Smith proposed openstack/glance master: Add image_set_property_atomic() helper https://review.opendev.org/737868 | 16:40 |
openstackgerrit | Dan Smith proposed openstack/glance master: WIP: Fix race condition in copy image operation https://review.opendev.org/737596 | 16:40 |
openstackgerrit | Dan Smith proposed openstack/glance master: WIP: FakeImage test fixture is not reflective of reality https://review.opendev.org/738241 | 16:40 |
*** priteau has quit IRC | 16:48 | |
mordred | dansmith: yeah - immediately before my message about it here | 17:01 |
dansmith | mordred: okay my devstack patch got kicked out so I figured that's why.. I rechecked the full chain so we can see if the properties end up right | 17:01 |
mordred | cool | 17:02 |
*** udesale has quit IRC | 17:39 | |
jokke | dansmith: here | 17:49 |
*** suryasingh has quit IRC | 17:49 | |
dansmith | jokke: hey, please tell me I'm mistaken on this: https://review.opendev.org/#/c/738241/1 | 17:49 |
jokke | wait, what's going on there | 17:52 |
dansmith | seems to me that the import tests there just started using FakeImage for the return value of image_repol.get() at some point because it was "close enough" | 17:54 |
dansmith | abhishekk had implemented the image_repo pass through of my new method by just doing image_id, but that didn't seem right since the rest of those methods use an actual image, so when I changed it to use an image I noticed the discrepancy | 17:58 |
jokke | ohh ... well there was some changes recently (like during ussuri cycle, I think) to the fixtures. need to go back and see what was actually done there, but it's possible that there's been some mixup | 18:02 |
jokke | there's the db fixture and then there is the domain fixture. Has their usage been mixed at some point | 18:04 |
dansmith | unless the api used to get db objects directly and only just now gets the domain, I'm not sure what would have changed recently | 18:06 |
dansmith | all my glance history is about a week old so it's hard to tell what is new and what is old | 18:07 |
dansmith | since other code in images uses image.image_id I thought it was more likely that this was just erroneously wrong and nothing in the code that uses that particular FakeImage noticed | 18:08 |
jokke | dansmith: nope, it's nohing new, that's why the domain fixture is in there. And you're correct looking the tests, the fake image that's been specified to be returned by the mocked repo is indeed adhering to the db fixture, not to the domain fixture | 18:08 |
jokke | I wonder where else this is mixed up | 18:09 |
dansmith | seems like a lot of tests make up their own objects to mock return values, so I can see why this is error-prone | 18:12 |
dansmith | five 'class FakeImage' definitions in glance/tests/ | 18:16 |
jokke | dansmith: also makes no sense to have the fixture and that object in first place. That's why the fixtures exists, right? | 18:24 |
dansmith | seems like the tests that use the object did so just because it was easier to mock image_repo.get() than to have it pull objects out of the fixture, but...yeah :) | 18:25 |
dansmith | mordred: looks better.. are these "owner specified" things supposed to be there? I would have expected only the hw_rng_model passed on the command line | 20:37 |
dansmith | https://zuul.opendev.org/t/openstack/build/aa54890eefb04a2ba2b1fffdd4e74974/log/job-output.txt#21292 | 20:37 |
dansmith | I guess I should look at a regular run | 20:37 |
dansmith | okay yeah I see those in another run, so I guess that's expected | 20:38 |
mordred | dansmith: yes - owner_specified is expected - sdk creates some checksums of its own to handle knowing how to not re-upload something when it doesn't need to | 20:44 |
* mordred has to run out for a sec - biab | 20:44 | |
dansmith | cool, I googled and found some patches from you that seemed to indicate that | 20:44 |
dansmith | sounds like a good plan to me :) | 20:44 |
*** tosky has quit IRC | 23:00 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!