| *** baojg has quit IRC | 00:02 | |
| *** baojg has joined #openstack-manila | 00:14 | |
| *** baojg has quit IRC | 00:14 | |
| *** baojg has joined #openstack-manila | 00:15 | |
| *** baojg has quit IRC | 00:15 | |
| *** baojg has joined #openstack-manila | 01:16 | |
| *** baojg has quit IRC | 01:24 | |
| zhongjun | bswartz: I replied vponomaryov's inline commits, waiting for vponomaryov's response. | 01:35 |
|---|---|---|
| *** openstackgerrit has joined #openstack-manila | 01:38 | |
| openstackgerrit | Merged openstack/manila-ui master: Imported Translations from Zanata https://review.openstack.org/492067 | 01:38 |
| openstackgerrit | zhongjun proposed openstack/manila master: Add API document for share type quota https://review.openstack.org/491711 | 01:43 |
| openstackgerrit | zhongjun proposed openstack/manila master: Fix quota error in batch delete shares https://review.openstack.org/489501 | 02:06 |
| *** ianychoi_ has joined #openstack-manila | 02:11 | |
| bswartz | zhongjun: you fixed the issues he pointed out? | 02:16 |
| bswartz | at this point we're not looking to cut RC1 until tomorrow morning at the earliest | 02:17 |
| *** ianychoi has quit IRC | 02:18 | |
| *** dsariel has quit IRC | 02:18 | |
| *** openstackgerrit has quit IRC | 02:21 | |
| *** baojg has joined #openstack-manila | 02:21 | |
| zhongjun | bswartz: I just modify the function name that he pointed out. | 02:21 |
| zhongjun | bswartz: not do anything else in code | 02:22 |
| *** baojg has quit IRC | 02:26 | |
| *** openstackgerrit has joined #openstack-manila | 02:29 | |
| openstackgerrit | Merged openstack/manila master: Add API document for share group quotas https://review.openstack.org/491700 | 02:29 |
| openstackgerrit | zhongjun proposed openstack/manila master: Add API document for share type quota https://review.openstack.org/491711 | 02:36 |
| bswartz | zhongjun: regarding the renamed "context" variable, it would be clearer if you used a name like admin_ctx to separate it from the other variable | 03:16 |
| bswartz | also, perhaps add a comment about why the original variable is retained | 03:16 |
| bswartz | I was confused for the same reason vponomaryov was | 03:16 |
| zhongjun | bswartz: yeah, it would be better to make it more meaningful | 03:18 |
| openstackgerrit | zhongjun proposed openstack/manila master: Fix quota error in batch delete shares https://review.openstack.org/489501 | 03:22 |
| zhongjun | bswartz: agree, done | 03:22 |
| *** baojg has joined #openstack-manila | 04:23 | |
| *** baojg has quit IRC | 04:30 | |
| *** dsariel has joined #openstack-manila | 04:31 | |
| *** markstur has quit IRC | 04:32 | |
| *** sapcc-bot has joined #openstack-manila | 05:33 | |
| *** markstur has joined #openstack-manila | 05:35 | |
| *** sapcc-bot1 has quit IRC | 05:37 | |
| *** markstur has quit IRC | 05:40 | |
| *** lpetrut has joined #openstack-manila | 05:52 | |
| openstackgerrit | Ben Swartzlander proposed openstack/manila master: Re-enable broken CG code in NetApp driver https://review.openstack.org/491877 | 05:56 |
| openstackgerrit | Ben Swartzlander proposed openstack/manila master: Re-enable broken CG code in NetApp driver https://review.openstack.org/491877 | 06:02 |
| *** baojg has joined #openstack-manila | 06:27 | |
| *** jprovazn has joined #openstack-manila | 06:30 | |
| *** baojg has quit IRC | 06:34 | |
| openstackgerrit | zhongjun proposed openstack/manila master: [Doc] Fix access rule description in api-ref https://review.openstack.org/492817 | 06:34 |
| *** lpetrut has quit IRC | 06:52 | |
| *** rraja has joined #openstack-manila | 07:17 | |
| openstackgerrit | zhongjun proposed openstack/manila master: [Doc] Fix access rule description in api-ref https://review.openstack.org/492817 | 07:26 |
| *** baojg has joined #openstack-manila | 07:30 | |
| *** baojg has quit IRC | 07:37 | |
| *** dsariel_ has joined #openstack-manila | 07:57 | |
| *** dsariel has quit IRC | 07:57 | |
| *** openstackgerrit has quit IRC | 08:02 | |
| *** openstackgerrit has joined #openstack-manila | 08:26 | |
| openstackgerrit | junbo.li proposed openstack/manila master: [Api-ref] Add supported protocol "MAPRFS" in doc https://review.openstack.org/492864 | 08:26 |
| openstackgerrit | yfzhao proposed openstack/manila master: Fix wrong links https://review.openstack.org/492879 | 08:48 |
| *** dsariel_ has quit IRC | 09:04 | |
| *** jprovazn has quit IRC | 09:22 | |
| *** baojg has joined #openstack-manila | 09:34 | |
| *** rejy has joined #openstack-manila | 09:35 | |
| *** ianychoi_ is now known as ianychoi | 09:36 | |
| *** baojg has quit IRC | 09:40 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/manila-ui master: Imported Translations from Zanata https://review.openstack.org/492932 | 09:57 |
| *** ociuhandu has quit IRC | 10:08 | |
| *** jprovazn has joined #openstack-manila | 10:14 | |
| *** rejy has quit IRC | 10:18 | |
| *** markstur has joined #openstack-manila | 10:31 | |
| *** markstur has quit IRC | 10:36 | |
| *** ganso has joined #openstack-manila | 10:45 | |
| *** jprovazn has quit IRC | 11:24 | |
| *** jprovazn has joined #openstack-manila | 11:32 | |
| *** jprovazn has quit IRC | 11:53 | |
| *** jprovazn has joined #openstack-manila | 12:09 | |
| *** jprovazn is now known as jprovazn_mtg | 12:09 | |
| zhongjun | vponomaryov: ping | 12:09 |
| *** jprovazn_mtg is now known as jprovazn | 12:10 | |
| zhongjun | vponomaryov: link: https://review.openstack.org/#/c/489501/2/manila/share/api.py@926 | 12:10 |
| zhongjun | vponomaryov: Are you around | 12:11 |
| vponomaryov | zhongjun: pong | 12:12 |
| zhongjun | vponomaryov: How to check whether the "release of real resources" on storage backend be deleted? | 12:13 |
| vponomaryov | zhongjun: how do you think? | 12:13 |
| zhongjun | vponomaryov: We could have this case: We still have share record in db and the real resources on storage backend has been deleted | 12:13 |
| zhongjun | vponomaryov: In this case, we can still change the share status to available and force delete it. It means the share record in db still work and we still don't need to -1 shares quota | 12:14 |
| zhongjun | vponomaryov: Deletion of DB record could be tied to "quota calculation". | 12:15 |
| vponomaryov | zhongjun: I still don't see arguments against my proposal to remove quota handling in "share-api" and move to "share-manager" | 12:18 |
| zhongjun | vponomaryov: We only control the db in manila. If the DB record has deleted, it means the manila not need to manage this share, and we need to -1 shares quota. Am I missing something? | 12:18 |
| *** jprovazn has quit IRC | 12:19 | |
| *** guyr-infinidat has quit IRC | 12:19 | |
| *** jprovazn has joined #openstack-manila | 12:19 | |
| *** guyr-infinidat has joined #openstack-manila | 12:19 | |
| zhongjun | vponomaryov: Because we have two places to delete share instance in db. The one is code line 926 in manila-api service, the another one is code line 924, it will send delete share instance to manila-share service. | 12:20 |
| *** gouthamr has joined #openstack-manila | 12:20 | |
| *** lpetrut has joined #openstack-manila | 12:22 | |
| zhongjun | vponomaryov: The case is: 1. create a share 2. the scheduler can not find a host, raise error 3. create a share status change to "error" and the share doesn't exist 'host'. 4 delete this share. Then we will go to code line 926 and we can never go to share-manager | 12:23 |
| vponomaryov | zhongjun: the main goal you have is to make sure quota calculated correctly, no more, no less. There is no real need to delete DB record exactly in share-api service. Especially, if we are talking about "share-manager"-specific resources. | 12:23 |
| vponomaryov | zhongjun: user consumes "real" resources, not DB records | 12:24 |
| vponomaryov | zhongjun: it means, if share was not even scheduled anywhere, it should not consume quota as it is only DB record of failed attempt to create share | 12:25 |
| vponomaryov | zhongjun: in manila-api we check, if share doesn't have "host" value then it haven't taken any real resources and we can just delete DB record in place. But if it has "host" value, then we dedicate such right to "share-manager" | 12:26 |
| vponomaryov | in this case manila-api services will not have "concurrency" | 12:28 |
| zhongjun | vponomaryov: But the actually is we cost the shares quota even if we doesn't have "host" value | 12:28 |
| *** xyang1 has joined #openstack-manila | 12:29 | |
| vponomaryov | bug | 12:29 |
| zhongjun | vponomaryov: haha | 12:29 |
| vponomaryov | ? | 12:29 |
| zhongjun | vponomaryov: It looks like we still have many bug in quotas | 12:30 |
| *** catintheroof has joined #openstack-manila | 12:30 | |
| zhongjun | vponomaryov: :) | 12:30 |
| vponomaryov | bugs everywhere | 12:30 |
| vponomaryov | it is question of "whether you are able to leave with them or not" | 12:30 |
| vponomaryov | it is very minor bug | 12:31 |
| vponomaryov | because you can just force-delete DB record without any problem | 12:31 |
| vponomaryov | so, we can leave with it | 12:31 |
| zhongjun | vponomaryov: yeah, we only handle +1 shares quota in share-api services. | 12:32 |
| zhongjun | vponomaryov: We doesn't handle all errors when we create a share. | 12:35 |
| zhongjun | vponomaryov: we can leave with what? | 12:41 |
| openstackgerrit | Merged openstack/manila-ui master: Imported Translations from Zanata https://review.openstack.org/492932 | 12:41 |
| vponomaryov | zhongjun: with bug that we consider "just DB record" as resource consumption | 12:42 |
| openstackgerrit | Jan Provaznik proposed openstack/manila master: [Doc] Add more description to user messages api-ref https://review.openstack.org/492987 | 12:43 |
| zhongjun | vponomaryov: We handle most of all quota operate in manila-api services. I think we need to discuss it with other manila members | 12:45 |
| zhongjun | vponomaryov: How to check the real resources be created? | 12:49 |
| zhongjun | vponomaryov: We could get error even if the real share has created in backend storage | 12:50 |
| zhongjun | bswartz: ping | 12:51 |
| *** dustins has joined #openstack-manila | 12:54 | |
| *** baojg has joined #openstack-manila | 13:03 | |
| *** MVenesio has joined #openstack-manila | 13:07 | |
| *** MVenesio has quit IRC | 13:13 | |
| *** MVenesio has joined #openstack-manila | 13:13 | |
| openstackgerrit | Merged openstack/manila master: [Api-ref] Add supported protocol "MAPRFS" in doc https://review.openstack.org/492864 | 13:33 |
| *** jprovazn has quit IRC | 13:41 | |
| *** jprovazn has joined #openstack-manila | 13:41 | |
| *** jmlowe_ has joined #openstack-manila | 13:47 | |
| *** jmlowe has quit IRC | 13:49 | |
| *** cknight has joined #openstack-manila | 13:54 | |
| *** rraja has quit IRC | 13:55 | |
| bswartz | hey zhongjun | 14:06 |
| bswartz | gouthamr vponomaryov: https://review.openstack.org/#/c/491877/ | 14:07 |
| *** markstur has joined #openstack-manila | 14:08 | |
| bswartz | zhongjun vponomaryov: are we not agreed on the approach to use for quota management? I'll admit that I'm not an expert on quotas, but tbarron had a specific proposal in Austin that I was okay with | 14:08 |
| vponomaryov | bswartz: there is no NetApp CI results so far | 14:08 |
| bswartz | perhaps we need to revisit the issue at PTG | 14:08 |
| bswartz | vponomaryov: doh | 14:08 |
| bswartz | I and all the tests and they passed, finally | 14:09 |
| bswartz | I'll go see why the CI system hasn't voted | 14:09 |
| *** rraja has joined #openstack-manila | 14:10 | |
| *** markstur has quit IRC | 14:12 | |
| bswartz | vponomaryov: hmm netapp CI seems to be not listening to event stream | 14:13 |
| zhongjun | bswartz: yes, vponomaryov think we need to remove most of all quota operate from manila-api service to manila-manager, in order to change "just DB record" as resource consumption to "real resource consumption" | 14:15 |
| bswartz | zhongjun: we're not going to fix the architectural issues in a bugfix | 14:15 |
| bswartz | we should solve the architectural issues in queens | 14:16 |
| bswartz | the question is whether this bugfix makes things better or not | 14:16 |
| zhongjun | vponomaryov: Do we still have problem about this bug except the architectural issues? | 14:18 |
| *** jprovazn has quit IRC | 14:19 | |
| *** markstur has joined #openstack-manila | 14:20 | |
| zhongjun | bswartz: I didn't see another problem about this bug fix except the architectural issues that vponomaryov pointed out | 14:20 |
| zhongjun | vponomaryov: Do I missing something? | 14:21 |
| *** markstur has quit IRC | 14:24 | |
| *** rraja has quit IRC | 14:25 | |
| bswartz | vponomaryov: I have no idea why the CI isn't posting comments, but here is the successful test run: http://paste.openstack.org/show/618189/ | 14:26 |
| bswartz | vponomaryov: looks like the CI system has developed a new bug >_< it's failing to pull patches from gerrit | 14:31 |
| bswartz | vponomaryov: we restarted zuul and it will vote in about an hour | 15:04 |
| *** markstur has joined #openstack-manila | 15:04 | |
| vponomaryov | bswartz: gouthamr has already provided logs, +2'ed | 15:05 |
| bswartz | vponomaryov: thx | 15:06 |
| bswartz | vponomaryov: what about zhongjun's bugfix? not worth merging in pike? | 15:06 |
| vponomaryov | bswartz: personally, I disagree with the approach there | 15:07 |
| bswartz | so you prefer to leave the bug in until we have a better approach to fix all the related issues? | 15:07 |
| vponomaryov | bswartz: bugfix is simple, not harder than current approach | 15:07 |
| vponomaryov | just different | 15:07 |
| *** jprovazn has joined #openstack-manila | 15:08 | |
| *** jprovazn has quit IRC | 15:08 | |
| vponomaryov | bswartz: I prefer 'proper', from my point of view, fix, good if in Pike | 15:09 |
| bswartz | well if this fix isn't the right one, then we won't fix it in pike | 15:10 |
| bswartz | we'll have to do some fixes in queens and backport them after the release | 15:10 |
| bswartz | that's probably a better approach in any case | 15:10 |
| bswartz | improving the quota system will be high on my priority list of queens | 15:11 |
| vponomaryov | bswartz: looks like I am the only one who have reviewed it, you may want to find some other couple of eye sfor it | 15:11 |
| bswartz | tbarron: ^ are you back online yet? | 15:11 |
| bswartz | vponomaryov: I'm okay with not merging it if the fix isn't correct | 15:11 |
| vponomaryov | bswartz: it is too loud to say it is improving | 15:11 |
| vponomaryov | bswartz: just moving small pile of code to other place | 15:12 |
| bswartz | I just wanted to give the patch a closer look since zhongjun did the work to get it fixed by the deadline | 15:12 |
| bswartz | I'm going to untarget the quota bugs from RC1 and tag the release when the NetApp bugfix clears the gate | 15:12 |
| bswartz | then we can start planning for queens work and hunting for bugs to fix in RC2 | 15:13 |
| bswartz | zhongjun: thanks for your work on the quota bug, looks like we'll have to -1 it for now | 15:13 |
| *** baojg_ has joined #openstack-manila | 15:15 | |
| zhongjun | It is okay, because we still have some things need to discuss about quotes problems | 15:16 |
| *** baojg has quit IRC | 15:16 | |
| zhongjun | bswartz: It would be better if we solve all quotes problems at the same time | 15:20 |
| *** ociuhandu has joined #openstack-manila | 15:34 | |
| *** ociuhandu has quit IRC | 16:10 | |
| *** lpetrut has quit IRC | 16:18 | |
| * bswartz curses gate-manila-tempest-minimal-dsvm-lvm-centos-7 | 16:20 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: Fix quota usages update deleting same share from several API endpoints https://review.openstack.org/493071 | 16:22 |
| vponomaryov | bswawrtz, zhongjun: ^ | 16:22 |
| vponomaryov | haven't tested it on lab, and haven't updated unit tests, it is for testing and review purposes | 16:23 |
| bswartz | vponomaryov: I'd like tbarron to look at that too -- I'm curious how it compares to what other projects are doing for quotas | 16:26 |
| *** jmlowe_ has quit IRC | 16:34 | |
| *** jmlowe has joined #openstack-manila | 16:56 | |
| openstackgerrit | Merged openstack/manila master: Fix wrong links https://review.openstack.org/492879 | 17:22 |
| *** lpetrut has joined #openstack-manila | 19:25 | |
| openstackgerrit | Ben Swartzlander proposed openstack/manila master: Re-enable broken CG code in NetApp driver https://review.openstack.org/491877 | 19:41 |
| *** lpetrut has quit IRC | 20:01 | |
| *** lpetrut has joined #openstack-manila | 20:02 | |
| *** gouthamr has quit IRC | 20:14 | |
| *** MVenesio has quit IRC | 20:14 | |
| *** catintheroof has quit IRC | 20:15 | |
| *** gouthamr has joined #openstack-manila | 20:32 | |
| *** jmlowe has quit IRC | 20:45 | |
| *** jmlowe has joined #openstack-manila | 20:47 | |
| *** cknight has quit IRC | 20:53 | |
| *** jmlowe has quit IRC | 20:56 | |
| *** dustins has quit IRC | 21:00 | |
| *** MVenesio has joined #openstack-manila | 21:33 | |
| *** baojg has joined #openstack-manila | 21:37 | |
| *** baojg_ has quit IRC | 21:37 | |
| openstackgerrit | Merged openstack/manila master: Re-enable broken CG code in NetApp driver https://review.openstack.org/491877 | 21:50 |
| *** xyang1 has quit IRC | 21:54 | |
| *** MVenesio_ has joined #openstack-manila | 22:18 | |
| *** MVenesio has quit IRC | 22:22 | |
| *** lpetrut has quit IRC | 22:23 | |
| *** ganso has quit IRC | 22:31 | |
| *** MVenesio_ has quit IRC | 22:51 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/manila master: Update reno for stable/pike https://review.openstack.org/493257 | 23:00 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!