| *** k_mouza has joined #openstack-glance | 01:19 | |
| *** k_mouza has quit IRC | 01:24 | |
| *** k_mouza has joined #openstack-glance | 01:49 | |
| *** k_mouza has quit IRC | 01:53 | |
| *** gyee has quit IRC | 02:34 | |
| *** k_mouza has joined #openstack-glance | 02:42 | |
| *** k_mouza has quit IRC | 02:47 | |
| *** k_mouza has joined #openstack-glance | 03:20 | |
| *** k_mouza has quit IRC | 03:25 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #openstack-glance | 04:33 | |
| *** takamatsu has quit IRC | 04:38 | |
| *** takamatsu has joined #openstack-glance | 04:38 | |
| *** ratailor has joined #openstack-glance | 05:11 | |
| *** udesale has joined #openstack-glance | 05:13 | |
| *** ralonsoh has joined #openstack-glance | 05:43 | |
| openstackgerrit | Abhishek Kekane proposed openstack/glance master: Change database migration version to victoria https://review.opendev.org/743914 | 06:20 |
|---|---|---|
| *** lpetrut has joined #openstack-glance | 06:22 | |
| *** Luzi has joined #openstack-glance | 06:27 | |
| *** amoralej|off is now known as amoralej | 06:55 | |
| *** bhagyashris is now known as bhagyashris|lunc | 07:28 | |
| *** rcernin has quit IRC | 07:32 | |
| *** nikparasyr has joined #openstack-glance | 07:38 | |
| *** rosmaita has quit IRC | 08:13 | |
| *** rosmaita has joined #openstack-glance | 08:24 | |
| *** rcernin has joined #openstack-glance | 08:28 | |
| *** rcernin has quit IRC | 08:34 | |
| *** k_mouza has joined #openstack-glance | 09:36 | |
| *** bhagyashris|lunc is now known as bhagyashris | 09:37 | |
| *** k_mouza has quit IRC | 10:13 | |
| *** k_mouza has joined #openstack-glance | 10:35 | |
| *** rcernin has joined #openstack-glance | 10:37 | |
| *** rcernin has quit IRC | 10:54 | |
| *** udesale_ has joined #openstack-glance | 11:54 | |
| *** udesale has quit IRC | 11:57 | |
| *** amoralej is now known as amoralej|lunch | 12:05 | |
| *** k_mouza has quit IRC | 12:24 | |
| *** m75abrams has joined #openstack-glance | 12:25 | |
| *** lpetrut_ has joined #openstack-glance | 12:33 | |
| *** lpetrut has quit IRC | 12:36 | |
| *** k_mouza has joined #openstack-glance | 12:39 | |
| *** k_mouza has quit IRC | 12:43 | |
| *** tkajinam has quit IRC | 12:47 | |
| *** m75abrams has quit IRC | 12:57 | |
| *** amoralej|lunch is now known as amoralej | 13:05 | |
| *** ratailor has quit IRC | 13:07 | |
| *** k_mouza has joined #openstack-glance | 13:16 | |
| *** k_mouza has quit IRC | 13:20 | |
| *** k_mouza has joined #openstack-glance | 13:26 | |
| *** k_mouza has quit IRC | 13:30 | |
| abhishekk | rosmaita, could you please put your insight on https://review.opendev.org/743914 | 13:40 |
| *** k_mouza has joined #openstack-glance | 13:41 | |
| openstackgerrit | Dan Smith proposed openstack/glance master: Flesh out FakeImage for extra_properties https://review.opendev.org/743839 | 13:41 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Add image_delete_property_atomic() helper https://review.opendev.org/743596 | 13:41 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Heartbeat the actual work of the task https://review.opendev.org/743426 | 13:41 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Update task message during import https://review.opendev.org/743427 | 13:41 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Implement time-limited import locking https://review.opendev.org/743597 | 13:41 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Add testing for _CompleteTask in api_image_import https://review.opendev.org/743593 | 13:44 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Add a test to make sure import get_flow() includes all the steps we expect https://review.opendev.org/743594 | 13:44 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Add tests for _ImportToStore.execute() https://review.opendev.org/743595 | 13:44 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Flesh out FakeImage for extra_properties https://review.opendev.org/743839 | 13:44 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Add image_delete_property_atomic() helper https://review.opendev.org/743596 | 13:44 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Heartbeat the actual work of the task https://review.opendev.org/743426 | 13:44 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Update task message during import https://review.opendev.org/743427 | 13:44 |
| openstackgerrit | Dan Smith proposed openstack/glance master: Implement time-limited import locking https://review.opendev.org/743597 | 13:44 |
| *** k_mouza has quit IRC | 13:45 | |
| *** k_mouza has joined #openstack-glance | 13:50 | |
| *** yoctozepto has quit IRC | 13:54 | |
| *** k_mouza has quit IRC | 13:54 | |
| *** yoctozepto has joined #openstack-glance | 13:55 | |
| *** k_mouza has joined #openstack-glance | 13:59 | |
| abhishekk | rosmaita, smcginnis, jokke weekly meeting at #openstack-meeting | 14:00 |
| *** k_mouza has quit IRC | 14:04 | |
| *** Luzi_ has joined #openstack-glance | 14:10 | |
| *** k_mouza has joined #openstack-glance | 14:10 | |
| *** Luzi has quit IRC | 14:11 | |
| *** k_mouza has quit IRC | 14:14 | |
| *** k_mouza has joined #openstack-glance | 14:15 | |
| openstackgerrit | Dan Smith proposed openstack/glance master: Implement time-limited import locking https://review.opendev.org/743597 | 14:17 |
| *** lpetrut_ has quit IRC | 14:19 | |
| *** k_mouza has quit IRC | 14:19 | |
| *** k_mouza has joined #openstack-glance | 14:22 | |
| *** k_mouza has quit IRC | 14:27 | |
| *** alistarle has joined #openstack-glance | 14:34 | |
| smcginnis | abhishekk: Looks like definitely not needed: https://review.opendev.org/#/c/701041/ | 14:38 |
| abhishekk | smcginnis, ack | 14:38 |
| *** bhagyashris is now known as bhagyashris|away | 14:47 | |
| jokke | alistarle: if you have time after the meeting I'm happy to have a chat about the rbd stuff | 14:54 |
| *** alistarle has quit IRC | 14:56 | |
| dansmith | guess not :) | 14:56 |
| jokke | I guess so | 14:56 |
| dansmith | rosmaita: I don't really know what "cinder multiple stores" means, but I assume you have a job for testing that in cinder itself right? | 15:00 |
| rosmaita | dansmith: not really | 15:00 |
| rosmaita | iirc | 15:00 |
| rosmaita | the situation is that | 15:00 |
| dansmith | for some reason? | 15:00 |
| rosmaita | glance cinder store can only use one volume-type | 15:01 |
| rosmaita | so if you want your glance users to use >1 volume type | 15:01 |
| dansmith | currently, yeah | 15:01 |
| dansmith | yeah I get what the spec is about | 15:01 |
| rosmaita | ok, so then it doesn't really affect multiple cinder backends | 15:01 |
| dansmith | okay, I thought there was also a problem with multiple cinder backends, not just the volume type thing | 15:02 |
| rosmaita | i guess what i'm saying is that there is no clean mapping to what glance_store is doing and cinder tests | 15:02 |
| dansmith | but certainly cinder has a job where multiple volume types are available right? | 15:03 |
| rosmaita | dansmith: not that i'm aware of, but i may have too simplistic a view of the situation | 15:03 |
| rosmaita | dansmith: i hope so | 15:03 |
| *** k_mouza has joined #openstack-glance | 15:04 | |
| dansmith | point is, maybe it would be good to make this glance spec include the "cinder backing job, and with multiple volume types" as a work item | 15:04 |
| dansmith | which, if you have a job for that already, we just need to inherit and configure glance accordingly | 15:04 |
| rosmaita | there is probably a tempest api test, it would be in wiht the volumeadmin tests because the default is you must be admin to create volume types | 15:05 |
| dansmith | and you can have multiple volume types on a single backend or you need multiple backends for that? | 15:07 |
| rosmaita | you can have multiple types on a single backend | 15:08 |
| dansmith | okay, | 15:08 |
| *** k_mouza has quit IRC | 15:08 | |
| dansmith | and doesn't devstack create one or more for you ahead of running any tests? | 15:08 |
| rosmaita | types or backends? | 15:09 |
| dansmith | types | 15:09 |
| abhishekk | dansmith, rosmaita | 15:09 |
| abhishekk | https://review.opendev.org/#/c/743800/2 | 15:09 |
| abhishekk | this patch I proposed in devstack | 15:10 |
| dansmith | cool | 15:10 |
| abhishekk | I guess it also creates multiple backends in cinder | 15:10 |
| dansmith | rosmaita: presumably you need at least one type (the default) to be defined right? | 15:10 |
| rosmaita | since train, there is a default that comes automatically with cinder | 15:10 |
| rosmaita | cleverly named __DEFAULT__ | 15:10 |
| rosmaita | but i think tempest creates its own so it knows they are there and what they are named | 15:11 |
| *** k_mouza has joined #openstack-glance | 15:12 | |
| dansmith | okay I would have expected devstack to do that, | 15:12 |
| dansmith | so that if you run devstack and not cinder, you can create volumes | 15:12 |
| rosmaita | devstack probably does something, but i don't know what | 15:13 |
| dansmith | abhishekk: are you going to create a job that depends on that devstack patch and configures it thusly? | 15:13 |
| abhishekk | dansmith, I have already created | 15:13 |
| abhishekk | https://review.opendev.org/742373 | 15:14 |
| dansmith | ah sweet! | 15:15 |
| dansmith | abhishekk: so those code changes are bugs you found by running it for real? | 15:15 |
| rosmaita | i think so, looks like an image was uploaded to glance to store in cinder before the cinder service was started | 15:16 |
| abhishekk | dansmith, nope, that is Poc for the specs which you are discussing with Brian | 15:16 |
| *** k_mouza has quit IRC | 15:16 | |
| dansmith | abhishekk: what I mean is, you have some code changes in with the change to add the job | 15:17 |
| dansmith | like this: https://review.opendev.org/#/c/742373/4/glance/api/authorization.py | 15:17 |
| abhishekk | rosmaita, I have moved the upload image function in devstack after service of cinder starts | 15:17 |
| rosmaita | yes, i saw that ... good job! | 15:17 |
| abhishekk | dansmith, yes | 15:17 |
| dansmith | abhishekk: okay i think I see.. this is the _actual_ change plus the job | 15:18 |
| abhishekk | dansmith, yes, | 15:19 |
| dansmith | gotcha | 15:19 |
| *** k_mouza has joined #openstack-glance | 15:21 | |
| abhishekk | dansmith, went through your comments on Introspect plugin fix, will address them soon | 15:21 |
| abhishekk | thank you for review | 15:21 |
| dansmith | abhishekk: cool | 15:21 |
| dansmith | abhishekk: by the way, if we were to snap the qcow header during streaming, we could assert that the thing they uploaded was actually a qcow, in the case where the format is restricted | 15:23 |
| dansmith | just another benefit if desired | 15:23 |
| abhishekk | dansmith, yes | 15:24 |
| *** gyee has joined #openstack-glance | 15:25 | |
| *** k_mouza has quit IRC | 15:25 | |
| * abhishekk going for dinner, be back in 45 mins | 15:26 | |
| *** k_mouza has joined #openstack-glance | 15:28 | |
| *** k_mouza has quit IRC | 15:32 | |
| *** k_mouza has joined #openstack-glance | 15:33 | |
| *** Luzi_ has quit IRC | 15:36 | |
| *** k_mouza has quit IRC | 15:37 | |
| openstackgerrit | Merged openstack/glance_store master: [Trivial]Add missing white space between words https://review.opendev.org/743474 | 15:43 |
| * abhishekk back from break | 15:58 | |
| openstackgerrit | Abhishek Kekane proposed openstack/glance master: Change database migration version to victoria https://review.opendev.org/743914 | 15:58 |
| *** udesale_ has quit IRC | 16:13 | |
| *** amoralej is now known as amoralej|off | 16:22 | |
| openstackgerrit | Erno Kuvaja proposed openstack/glance master: Fix active image when all uploads fail https://review.opendev.org/744010 | 16:23 |
| openstackgerrit | Erno Kuvaja proposed openstack/glance master: Fix active image without data https://review.opendev.org/744011 | 16:23 |
| openstackgerrit | Erno Kuvaja proposed openstack/glance master: Fix active image without data https://review.opendev.org/744011 | 16:29 |
| openstackgerrit | Erno Kuvaja proposed openstack/glance master: Fix active image when all uploads fail https://review.opendev.org/744010 | 16:52 |
| openstackgerrit | Erno Kuvaja proposed openstack/glance master: Fix active image without data https://review.opendev.org/744011 | 16:52 |
| dansmith | abhishekk: for example: https://pastebin.com/Rxuat4GA | 16:58 |
| dansmith | abhishekk: https://pastebin.com/zx5cLdYM | 16:58 |
| dansmith | no need to store the full image | 16:59 |
| abhishekk | dansmith, looking | 16:59 |
| dansmith | that only (needs to) store 32 bytes of the image to figure out the virtual size | 17:02 |
| abhishekk | dansmith, yeah | 17:02 |
| abhishekk | I was under assumption that it requires entire image in local disk | 17:03 |
| dansmith | and would let you enforce upload format restrictions too | 17:03 |
| dansmith | nope | 17:03 |
| abhishekk | yeah | 17:04 |
| abhishekk | SO there is no need to introspect plugin at all | 17:05 |
| dansmith | it would be ideal to do this in the stream everywhere, unless there's some format we need to handle otherwise | 17:05 |
| dansmith | what other formats does glance care about? vmdk and hdi or something? | 17:06 |
| dansmith | ah I found the list | 17:06 |
| abhishekk | glance do support various formats | 17:07 |
| abhishekk | yep, all those in the list | 17:07 |
| dansmith | well, it'd probably be worth checking them all to see if we can do the same, but for most of the composite formats the goal is the same, so I'm sure they have a usable header | 17:08 |
| dansmith | even grabbing the first 2k of the file, dropping it into a temp file and running qemu-img on it will probably do the right thing in a lot of cases | 17:08 |
| abhishekk | worth trying | 17:09 |
| dansmith | qemu-img doesn't claim to support vhd, only vhdx but it might be similar enough that it does.. it can't create them at least | 17:10 |
| openstackgerrit | Abhishek Kekane proposed openstack/glance-specs master: Make cinder driver compatible with multiple stores https://review.opendev.org/695152 | 17:13 |
| abhishekk | dansmith, ack | 17:14 |
| abhishekk | dansmith, also AFAIK I never come across any bug in glance which claims using images other than qcow2 or raw | 17:17 |
| dansmith | surely anyone that uses the hyperv or vmware driver (at least) uses those image types though right? | 17:17 |
| abhishekk | but yeah, we need to cover about all the disk-formats which we say we support | 17:17 |
| abhishekk | yes | 17:17 |
| dansmith | yeah | 17:18 |
| dansmith | vhd is quite a bit more complex to find the magic size value, but you still don't have to store the whole thing | 17:24 |
| abhishekk | ok | 17:25 |
| dansmith | we could also do our best to figure out the virtual size for upload (format checking is easy), and then on import, if we couldn't guess the size, then we run qemu-img as a last resord | 17:25 |
| abhishekk | If we were able to do it then the whole introspect spec will change | 17:26 |
| dansmith | I don't want to derail anything, but it does seem like it'd be a lot better if we can do it via stream to me | 17:27 |
| abhishekk | Or at the moment we can proceed with calculating virtual size for qcow and raw? | 17:27 |
| dansmith | meaning do it via stream for those formats, and via qemu-img for others? | 17:28 |
| abhishekk | yes, meaning for traditional image create we will do it via stream and in plugin we will do it via qemu-img | 17:29 |
| dansmith | oh okay sure | 17:29 |
| dansmith | best effort for regular upload | 17:29 |
| abhishekk | Ok, I will try to modify specs on that basis | 17:30 |
| dansmith | in all cases, we know the format they *claim* the image is supposed to be before they start either type of upload right? | 17:30 |
| abhishekk | yes, disk-format needs to be set before upload starts | 17:31 |
| dansmith | ack, so it's legit to not calculate virtual_size if they say it is format X and our inspection tells us that is wrong? | 17:33 |
| dansmith | assume you will pass a -f to qemu-img and fail if it refuses to parse the image? | 17:33 |
| abhishekk | yes, like that | 17:34 |
| dansmith | cool | 17:35 |
| openstackgerrit | Erno Kuvaja proposed openstack/python-glanceclient master: Fixes "stores" property added to the image https://review.opendev.org/744024 | 17:41 |
| *** ralonsoh has quit IRC | 18:00 | |
| *** nikparasyr has left #openstack-glance | 18:11 | |
| * abhishekk signing out for the day | 18:23 | |
| jokke | dansmith: just while you're testing stuff. ^^ client patch stops it to adding stores property when image-create-via-import is used with --stores ... once that is merged and released I have API patch for it ready | 18:25 |
| jokke | dansmith: I'm pretty sure you'll appreciate the "stores" not being shown without being actually populated | 18:26 |
| openstackgerrit | Erno Kuvaja proposed openstack/glance master: Add "stores" to disallowed properties https://review.opendev.org/744030 | 18:50 |
| openstackgerrit | Erno Kuvaja proposed openstack/glance master: Add "stores" to disallowed properties https://review.opendev.org/744030 | 19:30 |
| dansmith | jokke: ack | 21:12 |
| *** tkajinam has joined #openstack-glance | 22:05 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/glance master: [goal] Migrate glance jobs to focal https://review.opendev.org/743092 | 22:30 |
| openstackgerrit | Ghanshyam Mann proposed openstack/python-glanceclient master: [goal] Migrate testing to ubuntu focal https://review.opendev.org/744075 | 23:06 |
| *** rcernin has joined #openstack-glance | 23:13 | |
| *** rcernin has quit IRC | 23:14 | |
| *** rcernin has joined #openstack-glance | 23:14 | |
| openstackgerrit | Merged openstack/glance master: Change database migration version to victoria https://review.opendev.org/743914 | 23:37 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!