| openstackgerrit | Vipin Balachandran proposed openstack/cinder master: VMware: Storage policy support https://review.openstack.org/603938 | 00:38 |
|---|---|---|
| *** lixiaoy1 has joined #openstack-cinder | 00:41 | |
| openstackgerrit | zhufl proposed openstack/cinder master: services api ref: fix field enum value and add missing field https://review.openstack.org/603619 | 00:43 |
| *** hedvig_01 has quit IRC | 00:45 | |
| *** dhineshob has quit IRC | 00:46 | |
| jungleboyj | lbragstad: That is a good question. It works somehow as users can see their own volumes and admin can see all volumes. geguileo knows the database stuff better. Maybe he has insight ... | 00:46 |
| *** imacdonn has quit IRC | 00:49 | |
| *** cloudnull has quit IRC | 01:00 | |
| *** tezar has quit IRC | 01:00 | |
| *** tobias-urdin has quit IRC | 01:00 | |
| *** Guest58757 has joined #openstack-cinder | 01:03 | |
| *** mriedem_away has quit IRC | 01:08 | |
| *** Zer0Byte_ has quit IRC | 01:08 | |
| *** imacdonn has joined #openstack-cinder | 01:09 | |
| *** imacdonn has quit IRC | 01:14 | |
| *** imacdonn has joined #openstack-cinder | 01:15 | |
| *** yumiriam has quit IRC | 01:23 | |
| *** zhaochao has joined #openstack-cinder | 01:27 | |
| *** mschuppert has quit IRC | 01:29 | |
| *** erlon has joined #openstack-cinder | 01:35 | |
| lbragstad | jungleboyj hmm - ok... it looks like it's because is_admin is set on the context? | 01:40 |
| jungleboyj | lbragstad: Ah, that could be. | 01:41 |
| lbragstad | https://github.com/openstack/cinder/blob/446a9381ebbe2eb1315f52735cad24f7da0a5bcb/cinder/db/sqlalchemy/api.py#L1947 -> https://github.com/openstack/cinder/blob/446a9381ebbe2eb1315f52735cad24f7da0a5bcb/cinder/db/sqlalchemy/api.py#L1926 | 01:42 |
| lbragstad | https://github.com/openstack/cinder/blob/446a9381ebbe2eb1315f52735cad24f7da0a5bcb/cinder/db/sqlalchemy/api.py#L109-L114 | 01:44 |
| jungleboyj | The difference between the admin and non-admin one is if the volume_admin_metadata is returned. | 01:46 |
| jungleboyj | Which makes sense. | 01:46 |
| lbragstad | hmm | 01:46 |
| lbragstad | you mean the difference between how the queries are built? | 01:47 |
| *** _hemna has quit IRC | 01:47 | |
| jungleboyj | Right. | 01:48 |
| lbragstad | jungleboyj ahh | 01:48 |
| jungleboyj | Because the owner of a volume should be able to see the volume but not the admin portion. | 01:49 |
| *** _hemna has joined #openstack-cinder | 01:49 | |
| lbragstad | so maybe it's a problem in how i'm mocking the volume using the fakes? | 01:50 |
| lbragstad | https://review.openstack.org/#/c/602489/3/cinder/tests/unit/api/v3/test_volume_protection.py@75 | 01:50 |
| jungleboyj | Let me look here. | 01:50 |
| * jungleboyj wants to note that I am terrible at mocks | 01:51 | |
| lbragstad | i guess the big thing i was curious about was if a user that creates a volume can actually retrieve it | 01:51 |
| jungleboyj | lbragstad: Yes. | 01:51 |
| lbragstad | ok - then that points me in the right direction, if that's possible (and I assume things would be broken if that wasn't true) then i'm more than likely doing something wrong in the test | 01:52 |
| *** erlon has quit IRC | 01:52 | |
| lbragstad | i saw mocks used in other places for creating volumes to test the API, but i didn't go digging in the mock code all that much | 01:52 |
| jungleboyj | Yeah, looking at the test here. | 01:53 |
| jungleboyj | lbragstad: Oh, look at the next test. | 01:54 |
| jungleboyj | Wait, you wrote that one too though. | 01:55 |
| lbragstad | https://review.openstack.org/#/c/602489/3/cinder/tests/unit/api/v3/test_volume_protection.py@75 ? | 01:55 |
| lbragstad | yeah - both of those fail, but not at the actual assertion | 01:55 |
| jungleboyj | Ok. That makes sense. | 01:55 |
| lbragstad | they fail in the volume creation step | 01:55 |
| lbragstad | it looks like when a volume is created it's fetched, oto | 01:56 |
| lbragstad | too* | 01:56 |
| lbragstad | the get appears to be failing | 01:56 |
| lbragstad | because the context i'm building for the request doesn't have is_admin=True (which is normal?) | 01:56 |
| lbragstad | so it doesn't return the volume i think it should? | 01:57 |
| jungleboyj | Yeah. | 01:57 |
| jungleboyj | It has to be something with the volume not being created with the right user and project. | 01:58 |
| jungleboyj | We have some place you can easily create a fake colume but I can't find that right now. | 01:59 |
| jungleboyj | https://github.com/openstack/cinder/blob/master/cinder/tests/unit/fake_volume.py ? | 02:01 |
| lbragstad | i wonder if admin_metadata is being set on the object | 02:02 |
| lbragstad | by default or something? | 02:03 |
| jungleboyj | Well not in the fake_db_volume . | 02:03 |
| jungleboyj | https://github.com/openstack/cinder/blob/master/cinder/tests/unit/fake_volume.py#L34 | 02:03 |
| *** Dinesh_Bhor has joined #openstack-cinder | 02:04 | |
| *** hoonetorg has quit IRC | 02:04 | |
| lbragstad | yeah - i'm building the volume by just using a cinder.objects.Volume() | 02:05 |
| tommylikehu | hey lbragstad, basically you need to pass project id into volume object when creating. | 02:05 |
| lbragstad | "building" becuase I don't think i'm actually building it? it's just an object to give to the mock to return i think? | 02:05 |
| lbragstad | ahhh | 02:05 |
| jungleboyj | tommylikehu: Yay, someone smarter than me! ;-) | 02:05 |
| tommylikehu | lbragstad: line 75: https://review.openstack.org/#/c/602489/3/cinder/tests/unit/api/v3/test_volume_protection.py | 02:05 |
| tommylikehu | jungleboyj: lol | 02:06 |
| lbragstad | ok | 02:06 |
| jungleboyj | tommylikehu: Oh, that makes sense. | 02:06 |
| lbragstad | so the Volume object doesn't inspect the context to determine the project id? | 02:06 |
| tommylikehu | lbragstad: I guess so | 02:07 |
| lbragstad | interesting | 02:07 |
| jungleboyj | lbragstad: That, however would explain the problem. | 02:07 |
| lbragstad | yeah.. exactly | 02:08 |
| lbragstad | lol | 02:08 |
| lbragstad | if a user attempts to ask cinder for a volume that was created by someone else, and they aren't an admin, what does cinder do? | 02:09 |
| lbragstad | does cinder return a 404 or a 403? | 02:09 |
| jungleboyj | I think it is a 403. | 02:09 |
| jungleboyj | That or it may do 404 so as to not expose that the volume even exists. | 02:09 |
| lbragstad | sure... | 02:10 |
| jungleboyj | We generally try not to provide any more information than we have to. | 02:10 |
| openstackgerrit | Lance Bragstad proposed openstack/cinder master: Propose example volume protection tests https://review.openstack.org/602489 | 02:11 |
| lbragstad | nice - that ^ passes for me locally | 02:11 |
| lbragstad | thanks tommylikehu jungleboyj! | 02:11 |
| tommylikehu | lbragstad: kind of funny , we do have policy to ensure the request context is the owner but the question is we will retrieve the object from database before we check the policy, so it's 404 now:) | 02:12 |
| lbragstad | i left a note on the 404 vs 403 part though... | 02:12 |
| lbragstad | yeah - it's kind of a chicken and egg situation | 02:12 |
| tommylikehu | lbragstad: correct:) | 02:12 |
| lbragstad | i wouldn't be opposed to excepting a 404 to return a 403 for uniformity, because i think we hit things like this in keystone, too | 02:13 |
| lbragstad | but that's probably an API discussion for another time ;) | 02:14 |
| *** hoonetorg has joined #openstack-cinder | 02:17 | |
| jungleboyj | lbragstad: Yay for it passing! | 02:19 |
| lbragstad | ... maybe? | 02:19 |
| lbragstad | hmm | 02:20 |
| lbragstad | i can't make it fail.. | 02:20 |
| lbragstad | by changing the policy | 02:20 |
| jungleboyj | Ahip ir. | 02:20 |
| jungleboyj | Ship it | 02:20 |
| openstackgerrit | Merged openstack/os-brick stable/queens: FC Allow for multipath volumes with different LUNs https://review.openstack.org/594501 | 02:24 |
| lbragstad | i'll see if i can get something up that proves that this fails (or that the policy engine is working in tests) tomorrow | 02:26 |
| lbragstad | i appreciate the help! | 02:26 |
| jungleboyj | lbragstad: Thank you for continuing to work on it. | 02:26 |
| *** Dinesh_Bhor has quit IRC | 03:11 | |
| *** Dinesh_Bhor has joined #openstack-cinder | 03:12 | |
| *** caixiaoyu has joined #openstack-cinder | 03:19 | |
| *** caixiaoyu has joined #openstack-cinder | 03:20 | |
| *** caixiaoyu has quit IRC | 03:22 | |
| *** caixiaoyu has joined #openstack-cinder | 03:23 | |
| *** sapd1_ has quit IRC | 03:45 | |
| *** sapd1 has joined #openstack-cinder | 03:45 | |
| *** Dinesh_Bhor has quit IRC | 03:50 | |
| *** Dinesh_Bhor has joined #openstack-cinder | 04:46 | |
| *** Guest58757 is now known as cloudnull | 05:32 | |
| *** brinzhang has joined #openstack-cinder | 05:36 | |
| *** alexchadin has joined #openstack-cinder | 05:39 | |
| *** Luzi has joined #openstack-cinder | 05:44 | |
| tommylikehu | lbragstad: we have lots of similar testcases that proof our policy system works, but not for the volume show api, see one of these: https://github.com/openstack/cinder/blob/b669eeab36e1d179f3aa7cb36281adbda6fe78ca/cinder/tests/unit/api/contrib/test_hosts.py#L215 | 05:53 |
| *** Dinesh_Bhor has quit IRC | 06:08 | |
| *** alexchadin has quit IRC | 06:08 | |
| openstackgerrit | yenai proposed openstack/os-brick master: Improve docstrings https://review.openstack.org/603991 | 06:12 |
| *** Dinesh_Bhor has joined #openstack-cinder | 06:15 | |
| *** dpawlik_ has joined #openstack-cinder | 06:38 | |
| *** Dinesh_Bhor has quit IRC | 06:38 | |
| *** belmoreira has joined #openstack-cinder | 06:39 | |
| *** dpawlik has quit IRC | 06:39 | |
| *** Dinesh_Bhor has joined #openstack-cinder | 06:42 | |
| *** dpawlik_ has quit IRC | 06:45 | |
| *** dpawlik has joined #openstack-cinder | 06:45 | |
| *** gkadam has joined #openstack-cinder | 06:46 | |
| *** rcernin has quit IRC | 07:02 | |
| *** e0ne has joined #openstack-cinder | 07:16 | |
| openstackgerrit | yenai proposed openstack/os-brick master: Remove the split mistake https://review.openstack.org/604004 | 07:37 |
| *** Dinesh_Bhor has quit IRC | 07:39 | |
| *** Dinesh_Bhor has joined #openstack-cinder | 07:49 | |
| *** Dinesh_Bhor has quit IRC | 07:54 | |
| *** Dinesh_Bhor has joined #openstack-cinder | 08:01 | |
| *** zhongjun2_ has quit IRC | 08:04 | |
| *** hoangcx has quit IRC | 08:04 | |
| *** hoangcx has joined #openstack-cinder | 08:04 | |
| openstackgerrit | Merged openstack/cinder master: Fix typos in volume api ref doc https://review.openstack.org/603263 | 08:26 |
| openstackgerrit | zhufl proposed openstack/cinder master: Fix some inconsistencies in qos-specs api ref https://review.openstack.org/604020 | 08:55 |
| *** lixiaoy1 has quit IRC | 08:58 | |
| *** lixiaoy11 has joined #openstack-cinder | 08:58 | |
| *** Bhujay has joined #openstack-cinder | 09:00 | |
| *** Dinesh_Bhor has quit IRC | 09:01 | |
| *** lixiaoy11 has quit IRC | 09:03 | |
| *** Dinesh_Bhor has joined #openstack-cinder | 09:19 | |
| *** Tahvok has left #openstack-cinder | 09:19 | |
| *** tobias-urdin has joined #openstack-cinder | 09:35 | |
| openstackgerrit | hzliaoyuehua proposed openstack/cinder master: for test https://review.openstack.org/604025 | 09:35 |
| *** belmoreira has quit IRC | 09:38 | |
| *** moshele has joined #openstack-cinder | 09:38 | |
| openstackgerrit | yenai proposed openstack/os-brick master: Improve docstrings https://review.openstack.org/603991 | 09:40 |
| *** pcaruana has joined #openstack-cinder | 10:04 | |
| *** Dinesh_Bhor has quit IRC | 10:13 | |
| *** Dinesh_Bhor has joined #openstack-cinder | 10:15 | |
| *** Dinesh_Bhor has quit IRC | 10:16 | |
| *** alexchadin has joined #openstack-cinder | 10:27 | |
| *** alexchadin has quit IRC | 10:28 | |
| *** alexchadin has joined #openstack-cinder | 10:28 | |
| *** alexchadin has quit IRC | 10:33 | |
| *** abishop has joined #openstack-cinder | 10:51 | |
| *** ganso has joined #openstack-cinder | 10:54 | |
| *** belmoreira has joined #openstack-cinder | 10:57 | |
| *** erlon has joined #openstack-cinder | 11:07 | |
| openstackgerrit | Rajat Dhasmana proposed openstack/cinder master: Fix multiattach set to false after retype https://review.openstack.org/604040 | 11:08 |
| *** gnufied has joined #openstack-cinder | 11:08 | |
| *** pcaruana has quit IRC | 11:15 | |
| *** pcaruana has joined #openstack-cinder | 11:20 | |
| *** alexchadin has joined #openstack-cinder | 11:28 | |
| *** luizbag has joined #openstack-cinder | 11:29 | |
| *** pcaruana has quit IRC | 11:32 | |
| *** alexchadin has quit IRC | 11:33 | |
| *** gnufied has quit IRC | 11:34 | |
| *** pcaruana has joined #openstack-cinder | 11:39 | |
| openstackgerrit | hzliaoyuehua proposed openstack/cinder master: Fix TypeError when retrieve pools filtered by volume-type https://review.openstack.org/604067 | 11:47 |
| *** pcaruana has quit IRC | 11:50 | |
| *** gnufied has joined #openstack-cinder | 11:57 | |
| *** alexchadin has joined #openstack-cinder | 12:08 | |
| *** gnufied has quit IRC | 12:14 | |
| *** dustins has joined #openstack-cinder | 12:26 | |
| *** dustins has quit IRC | 12:31 | |
| *** tpsilva has joined #openstack-cinder | 12:34 | |
| *** dustins has joined #openstack-cinder | 12:35 | |
| *** gnufied has joined #openstack-cinder | 12:41 | |
| *** dustins has quit IRC | 12:43 | |
| *** dustins has joined #openstack-cinder | 12:43 | |
| *** Bhujay has quit IRC | 12:43 | |
| *** Bhujay has joined #openstack-cinder | 12:44 | |
| *** Bhujay has quit IRC | 12:45 | |
| *** Bhujay has joined #openstack-cinder | 12:45 | |
| *** gkadam has quit IRC | 12:46 | |
| *** alexchadin has quit IRC | 12:48 | |
| *** gkadam has joined #openstack-cinder | 12:48 | |
| *** alexchadin has joined #openstack-cinder | 12:48 | |
| *** gkadam has quit IRC | 12:48 | |
| *** alexchadin has quit IRC | 13:01 | |
| *** mriedem has joined #openstack-cinder | 13:06 | |
| *** finix has joined #openstack-cinder | 13:27 | |
| *** caixiaoyu has quit IRC | 13:28 | |
| *** jistr is now known as jistr|call | 13:32 | |
| *** alexchadin has joined #openstack-cinder | 13:46 | |
| *** brinzhang has quit IRC | 13:46 | |
| *** e0ne has quit IRC | 13:48 | |
| lbragstad | does cinder override policies in unit tests? | 14:11 |
| *** dpawlik has quit IRC | 14:11 | |
| *** moshele has quit IRC | 14:11 | |
| smcginnis | lbragstad: I think I remember seeing that somewhere. In meeting now, but if you can't find an answer, I will try digging later. | 14:12 |
| lbragstad | smcginnis no worries - i'll keep digging and see what i can find | 14:13 |
| openstackgerrit | Keith Berger proposed openstack/cinder stable/pike: VMware: Use vSphere template as snapshot format https://review.openstack.org/604103 | 14:15 |
| openstackgerrit | Sean McGinnis proposed openstack/cinder stable/ocata: ScaleIO: Prevent usage of unsafe volumes https://review.openstack.org/604105 | 14:17 |
| *** alexchadin has quit IRC | 14:21 | |
| *** alexchadin has joined #openstack-cinder | 14:25 | |
| lbragstad | nvm... i think i figured it out https://github.com/openstack/cinder/blob/master/cinder/test.py#L261-L270 | 14:28 |
| lbragstad | https://github.com/openstack/cinder/blob/master/cinder/tests/unit/policy.json | 14:29 |
| smcginnis | bingo | 14:29 |
| ericyoung | smcginnis, I added the new scaleio/vxflex maintainer to that most recent patch | 14:29 |
| smcginnis | Thanks ericyoung | 14:30 |
| ericyoung | I look in on each of them still but wanted to make sure that msabag was aware | 14:31 |
| openstackgerrit | Sean McGinnis proposed openstack/cinder stable/ocata: ScaleIO: Prevent usage of unsafe volumes https://review.openstack.org/604105 | 14:32 |
| *** jistr|call is now known as jistr | 14:34 | |
| *** Luzi has quit IRC | 14:37 | |
| *** KeithMnemonic has joined #openstack-cinder | 14:40 | |
| *** zhaochao has quit IRC | 14:41 | |
| *** dpawlik has joined #openstack-cinder | 14:41 | |
| KeithMnemonic | smcginnis: Sanity check please. This is in Rocky already AFAIK, but maybe I am reading it wrong. https://review.openstack.org/604103 | 14:41 |
| *** dpawlik has quit IRC | 14:45 | |
| smcginnis | KeithMnemonic: Yep, that merged in 2017. whoami-rajat must have just have only seen the patch against master and assumed it was recent. | 14:45 |
| KeithMnemonic | whoami-rajat: please remove you -1 | 14:46 |
| whoami-rajat | smcginnis: my bad, but IMO it still needs to be backported to queens first right? | 14:47 |
| smcginnis | whoami-rajat: It was committed in queens. | 14:49 |
| *** Bhujay has quit IRC | 14:51 | |
| whoami-rajat | smcginnis: ok, thanks for the insights. KeithMnemonic done | 14:52 |
| *** e0ne has joined #openstack-cinder | 14:56 | |
| openstackgerrit | Lance Bragstad proposed openstack/cinder master: DO NOT MERGE: Showcase policy protection with tests https://review.openstack.org/604115 | 14:56 |
| openstackgerrit | Lance Bragstad proposed openstack/cinder master: DO NOT MERGE: Add an example test for policy change https://review.openstack.org/604116 | 14:56 |
| lbragstad | jungleboyj tommylikehu smcginnis ^ | 14:57 |
| smcginnis | Thanks for digging in to that lbragstad | 14:57 |
| lbragstad | i was able to make the tests i wrote in https://review.openstack.org/#/c/602489/4 fail | 14:58 |
| lbragstad | more test coverage like that should make it so you can modify policies without being unaware of breaking users | 14:58 |
| KeithMnemonic | thanks whoami-rajat: , smcginnis: | 15:00 |
| jungleboyj | lbragstad: Cool. | 15:00 |
| *** etp has joined #openstack-cinder | 15:03 | |
| *** alexchadin has quit IRC | 15:04 | |
| lbragstad | i can propose some docs to oslo.policy that elaborates on how to unit test APIs without integration tests by using oslo.context if that is helpful | 15:04 |
| *** dklyle has joined #openstack-cinder | 15:04 | |
| smcginnis | lbragstad: That might be useful. I think this is probably a common thing that I would bet isn't well covered right now. | 15:07 |
| lbragstad | yeah - i would agree | 15:08 |
| *** finix has quit IRC | 15:10 | |
| *** finix has joined #openstack-cinder | 15:11 | |
| *** jiaopengju has quit IRC | 15:15 | |
| *** jiaopengju has joined #openstack-cinder | 15:15 | |
| openstackgerrit | Merged openstack/cinder master: Remove aoe-revalidate, aoe-discover, aoe-flush https://review.openstack.org/603231 | 15:18 |
| *** LiangFang has joined #openstack-cinder | 15:20 | |
| *** dave-mccowan has quit IRC | 15:25 | |
| *** Liang___ has joined #openstack-cinder | 15:27 | |
| *** Liang___ has quit IRC | 15:28 | |
| *** dave-mccowan has joined #openstack-cinder | 15:29 | |
| *** LiangFang has quit IRC | 15:31 | |
| *** yumiriam has joined #openstack-cinder | 15:44 | |
| *** uberjay has quit IRC | 15:50 | |
| *** uberjay has joined #openstack-cinder | 15:50 | |
| *** LiangFang has joined #openstack-cinder | 15:52 | |
| openstackgerrit | hzliaoyuehua proposed openstack/cinder master: Fix TypeError when retrieving pools filtered by volume-type https://review.openstack.org/604067 | 15:52 |
| LiangFang | Could anybody help to review:https://review.openstack.org/#/c/602571/ Specify image size to avoid glance backend resize when uploading volume to glance | 15:57 |
| jungleboyj | LiangFang: Looking. | 15:58 |
| LiangFang | thanks, this works well in my environment | 15:58 |
| *** finix has quit IRC | 16:00 | |
| openstackgerrit | Dhinesh Balasubramaniam proposed openstack/cinder master: Hedvig Cinder driver implementation https://review.openstack.org/433341 | 16:18 |
| *** keper7 has joined #openstack-cinder | 16:20 | |
| keper7 | Hi Cinder, I'm trying to import some Pure Storage volumes in and following their guide it says to use the cinder manageable-list command. My Cinder client is telling me that command does not exist, even though I see it in the documentation. ```cinder --version``` shows 4.0.1. Does anyone know which version supports this command? We really need it to succeed in a migration project. | 16:23 |
| smcginnis | keper7: It's a bit confusing because that was a microversion addition. So the client won't allow it unless you provide the version it was added or higher. | 16:24 |
| smcginnis | keper7: I think that was added in 3.8, but it appears our docs on that are not very clear. | 16:26 |
| smcginnis | keper7: But try "cinder --os-volume-api-version 3.8 manageable-list". | 16:26 |
| keper7 | Yes!! Thank you so much | 16:27 |
| smcginnis | Cool. No problem. | 16:27 |
| jungleboyj | :-) We really need to improve the documentation around micro-versions. | 16:27 |
| jungleboyj | I think I had that as a to-do coming out of the PTG. | 16:27 |
| smcginnis | Yet another example of why we need to change that client behavior like we talked about at the ptg. | 16:27 |
| jungleboyj | smcginnis: ++ | 16:28 |
| smcginnis | I think I said I would do it, but didn't see it in the #ACTION list. | 16:28 |
| smcginnis | Or someone did at least. | 16:28 |
| jungleboyj | Hmmm, let me look at that. | 16:28 |
| smcginnis | Wednesday morning IIRC. | 16:28 |
| jungleboyj | smcginnis: Ah, that was during the Review of the user Survey Results. I missed creating a ACTION on that one. | 16:35 |
| jungleboyj | Do you want to take looking at that one or want me to look at it? | 16:36 |
| smcginnis | I can take it. | 16:36 |
| jungleboyj | Ok. | 16:36 |
| jungleboyj | Ok. Action items updated. | 16:38 |
| *** Bhujay has joined #openstack-cinder | 16:40 | |
| *** e0ne has quit IRC | 16:41 | |
| *** Bhujay has quit IRC | 16:41 | |
| *** Bhujay has joined #openstack-cinder | 16:42 | |
| openstackgerrit | Sean McGinnis proposed openstack/python-cinderclient master: Default help output to include MV updates https://review.openstack.org/604160 | 16:44 |
| smcginnis | jungleboyj: ^ | 16:44 |
| smcginnis | +1, -2 --- #winning | 16:45 |
| *** hedvig_02 has joined #openstack-cinder | 16:45 | |
| openstackgerrit | Miriam Yumi proposed openstack/cinder-tempest-plugin master: Fix tempest_roles to run backup tests https://review.openstack.org/604161 | 16:47 |
| smcginnis | hedvig_02: No need for private messages. | 16:47 |
| smcginnis | I can't unblock you from running your tests. No one is blocking you from doing so. | 16:48 |
| hedvig_02 | @smcginnis: I wasn't sure if my user id was working for the openstack cinder channel | 16:48 |
| hedvig_02 | @smcginnis: "Abandoned. Restore once you've actually run local tests so this doesn't waste gate resources when it can easily be checked for failures. " We received this comment from you on our check in. Could you please explain the status abandoned? | 16:50 |
| smcginnis | hedvig_02: Yes. Despite repeated warnings to stop pushing up code without testing it locally, it keeps happening. So rather than letting that keep going through the gate and consuming resources there, I stopped it. | 16:52 |
| *** BLZbubba has joined #openstack-cinder | 16:53 | |
| smcginnis | hedvig_02: I thought I saw issues there, but rerunning tests locally looks OK, so I've restored it. | 16:53 |
| smcginnis | hedvig_02: But please do not push up future updates without running tests first. | 16:53 |
| hedvig_02 | @smcginnis: The tox runs are passing on our end. We are not sure why it's failing on openstack. We have done our preliminary tests before pushing to openstack gate resources | 16:53 |
| BLZbubba | hi guys i'm trying "openstack volume service list" and when i use tcpdump on the cinder node, it shows a connection to localhost:35357 instead of controller:5000 | 16:53 |
| BLZbubba | but there is nothing in /var/log/cinder/*log or /var/log/apache2/*log at all to say who is attempting this connection | 16:54 |
| smcginnis | hedvig_02: The last patches pep8 job failure could not have passed locally. | 16:54 |
| smcginnis | hedvig_02: Nor the several dozen other ones prior to that. | 16:54 |
| BLZbubba | how do i figure out who has the bogus keystone_authtoken section? | 16:54 |
| hedvig_02 | @smcginnis: I can show a screenshot which clearly displays tox -ep27 and tox -pep8 has passed. | 16:55 |
| smcginnis | hedvig_02: Yes, as I said, I tested this one and it is fine now. | 16:55 |
| hedvig_02 | okay | 16:55 |
| BLZbubba | i suspect something is up with wsgi but i have no idea how to debug that | 16:56 |
| whoami-rajat | BLZbubba: did you install openstack using devstack? | 16:57 |
| hedvig_02 | @smcginnis: I see the status Restored. Do we have to do run some command on us to resume the zuul run ? | 17:00 |
| smcginnis | hedvig_02: No, it should automatically go as if it was a new version posted. | 17:01 |
| hedvig_02 | @smcginnis: Thank you | 17:01 |
| smcginnis | hedvig_02: Thanks for checking on it. | 17:01 |
| BLZbubba | whoami-rajat: no | 17:01 |
| BLZbubba | i used https://docs.openstack.org/cinder/rocky/install/cinder-storage-install-ubuntu.html | 17:02 |
| BLZbubba | could this error in /var/log/apache2/cinder_error.log be the problem: The policy file policy.json could not be found. | 17:05 |
| whoami-rajat | BLZbubba: strange, can you share the output of openstack --debug volume service list? also your cinder.conf | 17:11 |
| whoami-rajat | BLZbubba: you can try and create a policy file using this sample https://docs.openstack.org/cinder/queens/sample_policy.html but that doesn't seem to be the problem here. | 17:13 |
| imacdonn | smcginnis jungleboyj: I've coded an implementation of get_manageable_volumes() for the ZFSSA iSCSI volume driver ... would we need a BP to get that added? Maybe it could be considered a bug that we have manage/unmanage-volume but not get_manageable_volumes :) | 17:13 |
| *** gkadam has joined #openstack-cinder | 17:14 | |
| jungleboyj | imacdonn: So, the offical person in me says it needs a BP. The person who is moving to Storyboard says don't worry about it. :-) | 17:15 |
| imacdonn | heh | 17:15 |
| jungleboyj | Don't worry about the BP. There is a new process coming. | 17:15 |
| jungleboyj | If it has a release note, that is what I care about. Creating a bug for it wouldn't be all bad. | 17:16 |
| jungleboyj | Because yeah, that should have probably been there. | 17:16 |
| BLZbubba | "GET /v2/os-services HTTP/1.1" 404 112 "-" "python-cinderclient" | 17:16 |
| imacdonn | I didn't do the manage/unmanage implementation (but I'm fixing a couple of issues with it) ... I'll look for a RN | 17:17 |
| imacdonn | speaking of fixing those issues, this one (doc only) is ready for approval: https://review.openstack.org/#/c/603936/ | 17:18 |
| imacdonn | the other one (code fix) got hung up in the zuul illness yesterday .. I'm waiting for the backlog to clear a bit before retrying it | 17:18 |
| jungleboyj | Just need to put a Release Note in with the change you are creating now imacdonn | 17:21 |
| imacdonn | right, can do that | 17:21 |
| jungleboyj | imacdonn: Thank you sir. | 17:22 |
| imacdonn | thank you :) | 17:22 |
| BLZbubba | when i run cinder-wsgi --port=8776 instead of apache, it also says "GET /v2/os-services HTTP/1.1" 404 112 | 17:24 |
| BLZbubba | could it mean that the volume service isn't registering properly or something like that? | 17:24 |
| openstackgerrit | Merged openstack/cinder master: [Optimize] Use OVO when retrieve volume object https://review.openstack.org/603552 | 17:33 |
| openstackgerrit | Merged openstack/cinder master: Add microversion history to api-ref https://review.openstack.org/603486 | 17:33 |
| *** Bhujay has quit IRC | 17:34 | |
| keper7 | Hey guys is cinder manageable-list supposed to work with Ceph? Or just "real" storage appliances like Pure Storage? I'm trying to test it in pre-prod with Ceph Luminous and its not throwing an error, but returning nothing. Wondering if this is expected behaviour or maybe I have something configured wrong? | 17:34 |
| jungleboyj | keper7: I don't think Ceph has that implemented yet. I feel like we talked about that last week. | 17:34 |
| imacdonn | It's implemnented in the rbd driver, but I dunno if it works | 17:35 |
| imacdonn | I plagiarised it quite a bit for my implementation ;) | 17:35 |
| keper7 | Ok, so not much point in my fighting with it right now. I'll wait and use the one with Pure, I'm told it is tested and works. Thanks again! | 17:36 |
| jungleboyj | imacdonn: Ruh roh. | 17:36 |
| imacdonn | https://github.com/openstack/cinder/blob/master/cinder/volume/drivers/rbd.py#L1554-L1597 | 17:36 |
| jungleboyj | imacdonn: So, they don't have unmanage. | 17:36 |
| imacdonn | yeah, I saw that | 17:36 |
| jungleboyj | keper7: Is it possible that your volumes are in-use? | 17:37 |
| jungleboyj | It won't show up in your list if there is more than one watcher on a volume as it thinks it is in use by a client. | 17:38 |
| imacdonn | if that's the case, it should just list with with safe_to_manage=False .... if it works ;) | 17:38 |
| jungleboyj | imacdonn: Oh, you are right. | 17:38 |
| jungleboyj | Hmm, I need to play with that on one of my clusters. | 17:39 |
| keper7 | rbd create --size 5 volumes/manageTest ; cinder --os-volume-api-version 3.8 manageable-list $pool | 17:39 |
| keper7 | returns an empty table | 17:39 |
| keper7 | where pool came from ```cinder get-pools``` for my rbd driver | 17:41 |
| jungleboyj | Hmmm. That isn't good. | 17:42 |
| *** dave-mccowan has quit IRC | 17:44 | |
| keper7 | I'm on Pike, could it not be implemented yet in my version? | 17:44 |
| imacdonn | Yes, it looks like it got implemented in Queens | 17:45 |
| jungleboyj | keper7: Ah, good catch. *phew* | 17:45 |
| keper7 | ah ok, well glad that adds up then. | 17:46 |
| *** dave-mccowan has joined #openstack-cinder | 17:46 | |
| keper7 | @imacdonn, for reference, how would you use safe_to_manage=False? I dont see it as an argument | 17:47 |
| imacdonn | keper7: it's decided by the driver | 17:48 |
| imacdonn | see https://docs.openstack.org/cinder/latest/contributor/drivers.html | 17:48 |
| jungleboyj | keper7: Right. If it is an image that can't be managed for whatever reason that is what you will see. | 17:48 |
| imacdonn | "safe_to_manage (boolean): Whether or not this volume is safe to manage according to the storage backend. For example, is the volume in use or invalid for any reason." | 17:48 |
| keper7 | An ok, I thought you meant I had to pass that to the manageable-list command to be able to see them. Thanks | 17:49 |
| imacdonn | np | 17:49 |
| BLZbubba | whoami-rajat: http://185.56.20.124/cinder.conf | 18:06 |
| BLZbubba | what's weird is that cinder-volume seems to be registering the lvm properly but any request to cinder-wsgi gives a 404 | 18:06 |
| BLZbubba | but don't they use the same cinder.conf? | 18:06 |
| whoami-rajat | BLZbubba: did you try using curl to hit the API if it succeeds, it maybe a problem at the openstack client side. | 18:09 |
| whoami-rajat | the conf seems fine | 18:09 |
| BLZbubba | here's the output from python2-cinder --debug service-list: http://185.56.20.124/service_error.txt | 18:10 |
| *** e0ne has joined #openstack-cinder | 18:10 | |
| BLZbubba | let me figure out how to do the tokens with curl, curl gives me a 401 currently | 18:10 |
| openstackgerrit | Rajat Dhasmana proposed openstack/cinder master: Fix multiattach set to false after retype https://review.openstack.org/604040 | 18:11 |
| BLZbubba | i wish i could see WTF cinder-wsgi is doing, it is currently just saying 404 even with debug=True | 18:11 |
| whoami-rajat | openstack token issue seems to generate valid authorized tokens | 18:12 |
| BLZbubba | ok cool it just says 404 still | 18:14 |
| BLZbubba | i think cinder-wsgi is jacked somehow | 18:14 |
| imacdonn | you should be able to GET http://controller:8776/ without a keystone token | 18:15 |
| imacdonn | (it should show available API versions) | 18:16 |
| BLZbubba | well this is on the cinder node | 18:17 |
| BLZbubba | hmm does that endpoint need to be on the controller? | 18:17 |
| BLZbubba | maybe that's where i'm going wrong | 18:17 |
| imacdonn | wherever cinder-api is running | 18:18 |
| BLZbubba | cinder-api and cinder-wsgi are the same thing, right? | 18:19 |
| BLZbubba | at least, cinder-wsgi seems to be the only important file in the cinder-api package | 18:20 |
| imacdonn | I think cinder-wsgi is intended to be run in something like uwsgi or mod_wsgi, whereas cinder-api is the eventlet-based version | 18:21 |
| BLZbubba | hmm, on ubuntu 18.04 the cinder-api package only contains some docs and /etc/apache2/conf-available/cinder-wsgi.conf | 18:22 |
| imacdonn | yeah, so that'd be for mod_wsgi | 18:22 |
| BLZbubba | the real cinder-wsgi executable is from the python-cinder package | 18:23 |
| imacdonn | so are you running cinder-wsgi manually, or via httpd ? | 18:25 |
| BLZbubba | i tried it both ways | 18:27 |
| BLZbubba | always gives a 404 on any uri other than / and /v3/ | 18:27 |
| imacdonn | does / show API versions ? | 18:31 |
| *** dhineshob has joined #openstack-cinder | 18:33 | |
| BLZbubba | it appears so, there is application/vnd.openstack.volume+json;version=3 and also a version=2 | 18:34 |
| *** LiangFang has quit IRC | 18:47 | |
| openstackgerrit | Merged openstack/cinder master: Fix typo https://review.openstack.org/597506 | 18:48 |
| BLZbubba | ugh this is just weird | 18:50 |
| BLZbubba | the apache log says "Loaded extension: os-services register /usr/lib/python2.7/dist-packages/cinder/api/extensions.py:140" | 18:53 |
| BLZbubba | does that mean it should stop giving a 404 on /os-services or is there another step or two? | 18:57 |
| openstackgerrit | Merged openstack/cinder master: trival: Fix some spacing issues https://review.openstack.org/597540 | 19:00 |
| openstackgerrit | Merged openstack/cinder master: Dell PS Driver moves to maintenance mode https://review.openstack.org/603441 | 19:01 |
| openstackgerrit | Merged openstack/cinder master: Remove os-image-create API extension https://review.openstack.org/568528 | 19:01 |
| *** e0ne has quit IRC | 19:03 | |
| BLZbubba | should the / uri give a 300 response? | 19:09 |
| *** dpawlik has joined #openstack-cinder | 19:12 | |
| imacdonn | BLZbubba: what are you actually trying to accomplish ? | 19:12 |
| *** dpawlik has quit IRC | 19:16 | |
| BLZbubba | imacdonn trying to make cinder-wsgi work for uris other than / /v2/ and /v3/ | 19:16 |
| BLZbubba | i get a 404 on everything else | 19:17 |
| imacdonn | BLZbubba: But ... are you trying to install a released version of some distro of OpenStack, or make a development environment, or ... ?? | 19:17 |
| BLZbubba | it is rocky from the standard ubuntu repo, using this guide: https://docs.openstack.org/cinder/rocky/install/cinder-controller-install-ubuntu.html | 19:18 |
| BLZbubba | cinder-volume seems to be reporting the correct lvm info | 19:18 |
| BLZbubba | the debug logs claim that things like os-services are initializing | 19:19 |
| imacdonn | I don't know if there's a forum for help with the Ubuntu distro, but it seems you need that ... basic install/configure/get-started help ... | 19:19 |
| imacdonn | I'm not familiar with the way things are done in the Ubuntu distro, so can only guess | 19:20 |
| jungleboyj | BLZbubba: You didn't actually make the endpoint http://controller:8776/v2/%\(project_id\)s did you? | 19:20 |
| jungleboyj | I saw you refernece that earlier. 'Controller' should be replaced with the system where CInder is running. | 19:20 |
| BLZbubba | i think this is ok, i'm not even using the endpoint really, just curl to port 8776 directly | 19:21 |
| imacdonn | that is using the endpoint :) but you're not getting it from the keystone catalog | 19:21 |
| *** dustins has quit IRC | 19:22 | |
| BLZbubba | hehe yeah | 19:22 |
| jungleboyj | Ok. | 19:22 |
| BLZbubba | i'm a n00b to wsgi, i'm trying to look in the code to figure out how the actual /v3/os-services uri is registered | 19:23 |
| imacdonn | you're missing the project ID part | 19:25 |
| imacdonn | you should be GET'ing something like '/v3/77e9ba96c4824317afc6c016725d5d7f/os-services' where "77e9ba96c4824317afc6c016725d5d7f" is the project ID associated with your keystone token | 19:26 |
| imacdonn | what does this say? openstack endpoint list --service volumev2 --interface public --format value --column URL | 19:28 |
| imacdonn | (or "volumev3") | 19:28 |
| BLZbubba | the correct url: http://os-cinder:8776/v2/ | 19:29 |
| imacdonn | that is not the correct URL | 19:29 |
| imacdonn | it should be "http://os-cinder:8776/v2/%(tenant_id)s" | 19:29 |
| BLZbubba | ah ok | 19:29 |
| imacdonn | er make that "project_id", not "tenant_id", although I think either will work | 19:30 |
| *** e0ne has joined #openstack-cinder | 19:30 | |
| BLZbubba | hehe i was just going to ask that | 19:31 |
| *** e0ne has quit IRC | 19:34 | |
| imacdonn | I need to fix that in my own deployment playbook | 19:34 |
| *** belmoreira has quit IRC | 19:35 | |
| BLZbubba | hmm should i just skip volumev2 from now on? | 19:36 |
| imacdonn | it's supposedly deprecated, but it still gets used by default in some places - e.g. openstacksdk/openstackclient | 19:37 |
| imacdonn | story for that: https://storyboard.openstack.org/#!/story/2003691 | 19:37 |
| BLZbubba | awesome, service list works now! | 19:39 |
| BLZbubba | you guys rock | 19:39 |
| imacdonn | \o/ | 19:39 |
| BLZbubba | _\m/ | 19:39 |
| BLZbubba | now the big question: will it magically work in horizon | 19:39 |
| openstackgerrit | Merged openstack/cinder master: Fix pylint warnings for "unnecessary not" https://review.openstack.org/601312 | 19:40 |
| *** e0ne has joined #openstack-cinder | 19:43 | |
| BLZbubba | volume attach worked on the first try. | 19:53 |
| BLZbubba | that is bad luck :P | 19:53 |
| imacdonn | hah | 19:53 |
| imacdonn | "Wait - that worked? That can't be right" | 19:54 |
| BLZbubba | ok next i need to get the nvmet version working. that is practically guaranteed not to work the first time | 20:04 |
| *** erlon has quit IRC | 20:07 | |
| *** gkadam has quit IRC | 20:10 | |
| *** dustins has joined #openstack-cinder | 20:12 | |
| *** luizbag has quit IRC | 20:18 | |
| openstackgerrit | Merged openstack/os-brick master: Improve docstrings https://review.openstack.org/603991 | 20:28 |
| *** keper7 has quit IRC | 20:39 | |
| *** dustins has quit IRC | 21:31 | |
| *** e0ne has quit IRC | 21:36 | |
| *** nikeshm has joined #openstack-cinder | 21:50 | |
| nikeshm | any date on deadline for submission of new driver for S release | 21:51 |
| nikeshm | any date on approval of bp for new driver for S release | 21:52 |
| openstackgerrit | iain MacDonnell proposed openstack/cinder master: ZFSSA iSCSI implement get_manageable_volumes() https://review.openstack.org/604230 | 21:57 |
| woojay | nikeshm: I think it will be similar to Manila timeline around early January for driver submission deadline. https://releases.openstack.org/stein/schedule.html | 21:58 |
| *** hedvig_02 has quit IRC | 22:07 | |
| openstackgerrit | Merged openstack/cinder stable/pike: ScaleIO: Prevent usage of unsafe volumes https://review.openstack.org/601681 | 22:10 |
| openstackgerrit | Merged openstack/cinder stable/queens: Add missing 'target_obj' when perform policy check https://review.openstack.org/593675 | 22:10 |
| openstackgerrit | Merged openstack/cinder stable/queens: Update delete group snapshot API exception handling https://review.openstack.org/593201 | 22:10 |
| KeithMnemonic | jungleboyj: Hi again ;-0 https://review.openstack.org/#/c/604103/ | 22:15 |
| *** ganso has quit IRC | 22:17 | |
| jungleboyj | KeithMnemonic: Will take a look later. | 22:17 |
| KeithMnemonic | thanks! | 22:18 |
| *** spsurya has quit IRC | 22:48 | |
| *** xyang has quit IRC | 22:48 | |
| *** david-lyle has joined #openstack-cinder | 22:49 | |
| *** spsurya has joined #openstack-cinder | 22:50 | |
| *** dklyle has quit IRC | 22:51 | |
| *** rcernin has joined #openstack-cinder | 22:53 | |
| *** mriedem is now known as mriedem_away | 23:10 | |
| *** dpawlik has joined #openstack-cinder | 23:12 | |
| imacdonn | any core team other than jungleboyj and smcginnis, please review https://review.openstack.org/#/c/593955/ | 23:15 |
| *** erlon has joined #openstack-cinder | 23:16 | |
| *** dpawlik has quit IRC | 23:16 | |
| openstackgerrit | Dhinesh Balasubramaniam proposed openstack/cinder master: Hedvig Cinder driver implementation https://review.openstack.org/433341 | 23:18 |
| openstackgerrit | Merged openstack/cinder stable/queens: Add user messages for extend volume operation https://review.openstack.org/593202 | 23:19 |
| *** brinzhang has joined #openstack-cinder | 23:26 | |
| *** andreykurilin has quit IRC | 23:34 | |
| *** andreykurilin has joined #openstack-cinder | 23:35 | |
| *** rcernin has quit IRC | 23:36 | |
| *** rcernin has joined #openstack-cinder | 23:36 | |
| *** dpawlik has joined #openstack-cinder | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!