| *** gouthamr_ has joined #openstack-manila | 01:05 | |
| *** gouthamr has quit IRC | 01:07 | |
| *** baojg_ has quit IRC | 01:10 | |
| openstackgerrit | Merged openstack/puppet-manila master: Add "host" parameter for manila.conf https://review.openstack.org/497481 | 01:23 |
|---|---|---|
| zhongjun | gouthamr_: ping | 01:27 |
| gouthamr_ | zhongjun: pong | 01:27 |
| zhongjun | gouthamr_: We can not get all shares by non-admin even if we change share_instance:index policy from ‘admin_api’ to non limit. Because we added | 01:27 |
| zhongjun | @require_admin_context in share_instances_get_all db function. Do you know any | 01:27 |
| zhongjun | historical reasons? | 01:27 |
| *** gouthamr_ is now known as gouthamr | 01:27 | |
| bswartz | zhongjun: we never test the nonstandard policies | 01:28 |
| gouthamr | zhongjun: sounds like a bug | 01:28 |
| bswartz | I would count it as a bug | 01:28 |
| zhongjun | gouthamr: I tested it when I write the code about "policy in code" feature, it will raise “ERROR: User does not have admin privileges.” | 01:29 |
| gouthamr | We could unit test policies.. | 01:29 |
| bswartz | gouthamr: if you can think of a scalable way to do so | 01:30 |
| bswartz | IMO the matrix of options is too large | 01:30 |
| gouthamr | bswartz: true.. but with zhongjun's changes, there'll be a default policy in code, we could specify a non-default configuration to test | 01:32 |
| zhongjun | bswartz: We added @require_admin_context in many places : https://github.com/openstack/manila/blob/master/manila/db/sqlalchemy/api.py#L372 I thought it could have a original reason | 01:32 |
| gouthamr | zhongjun: i used to see the pattern with deletes, most calls elevate privileges to delete the row | 01:33 |
| bswartz | I would be mostly concerned with testing the dangerous cases | 01:33 |
| gouthamr | zhongjun: wait-a-minute, that's many GET methods | 01:33 |
| bswartz | it's way worse if a non-admin can do admin things than if an admin can't do something | 01:34 |
| gouthamr | yes.. i think we've been mindful about the first part.. kinda why we split admin and regular api tests apart | 01:35 |
| zhongjun | bswartz: I changed "share:get_all" policy from ‘admin_api’ to 'default', it is not nonstandard policies. | 01:36 |
| gouthamr | zhongjun: would be a bug worth fixing.. we have the policy check in the service API as well... https://github.com/openstack/manila/blob/master/manila/api/v2/services.py#L35 | 01:37 |
| bswartz | zhongjun: as long as the policy is properly enforced (and we have tests to ensure that) then I'm happy | 01:39 |
| bswartz | it's entirely possible that there's some code in there that predates the policy.json mechanism and effectively hard codes the policy in some places | 01:40 |
| zhongjun | gouthamr, bswartz: yeah, it look like a bug. but I could try to find the original code modification, why they added this. | 01:41 |
| zhongjun | bswartz: yes, It could be. | 01:42 |
| zhongjun | bswartz: I didn't find it in nova, it could be changed from cinder | 01:43 |
| tommylikehu | maybe it could remind the administrator he/she had configured the wrong policy | 01:43 |
| zhongjun | tommylikehu: yes, It could be. I just think of the case: If the admin give a policy to the user to see all of the shares, is it a reasonable requirement? | 01:46 |
| *** markstur has quit IRC | 01:47 | |
| zhongjun | gouthamr, bswartz: In nova, it also added the admin check in db level after the policy check in API. But it didn't raise the error if the "xxx:get_all" policy changed from ‘admin_api’ to 'default'. It just add the appropriate filter(project, user) in db api level. https://github.com/openstack/nova/blob/16.0.0.0b2/nova/db/sqlalchemy/api.py#L2051 | 02:26 |
| zhongjun | gouthamr: yeah, we can not let a non-admin do admin things, it will be enforced in code | 02:28 |
| zhongjun | gouthamr: s / share:get_all / share_instance:index | 02:49 |
| *** gcb has joined #openstack-manila | 02:50 | |
| *** markstur has joined #openstack-manila | 02:53 | |
| *** markstur has quit IRC | 02:57 | |
| *** gouthamr has quit IRC | 02:59 | |
| *** baojg has joined #openstack-manila | 03:14 | |
| *** baojg has quit IRC | 03:19 | |
| *** markstur has joined #openstack-manila | 03:26 | |
| *** hoonetorg has quit IRC | 04:11 | |
| *** hoonetorg has joined #openstack-manila | 04:23 | |
| *** markstur has quit IRC | 04:24 | |
| *** markstur has joined #openstack-manila | 04:25 | |
| *** lpetrut has joined #openstack-manila | 04:26 | |
| *** markstur has quit IRC | 04:30 | |
| *** markstur has joined #openstack-manila | 04:46 | |
| *** rejy has joined #openstack-manila | 04:50 | |
| *** markstur has quit IRC | 04:50 | |
| *** lpetrut has quit IRC | 04:56 | |
| *** rejy has quit IRC | 05:22 | |
| *** markstur has joined #openstack-manila | 05:28 | |
| *** markstur has quit IRC | 05:32 | |
| *** carthaca_1 has quit IRC | 05:58 | |
| *** carthaca_ has joined #openstack-manila | 05:58 | |
| *** zengyingzhe has quit IRC | 06:14 | |
| *** zengyingzhe has joined #openstack-manila | 06:14 | |
| *** pcaruana has joined #openstack-manila | 06:39 | |
| *** raissa has quit IRC | 06:46 | |
| *** jprovazn has joined #openstack-manila | 07:10 | |
| *** markstur has joined #openstack-manila | 07:46 | |
| *** td has joined #openstack-manila | 07:49 | |
| *** markstur has quit IRC | 07:51 | |
| *** dsariel has joined #openstack-manila | 08:14 | |
| *** raissa has joined #openstack-manila | 08:20 | |
| *** markstur has joined #openstack-manila | 08:23 | |
| *** markstur has quit IRC | 08:28 | |
| *** markstur has joined #openstack-manila | 09:00 | |
| *** markstur has quit IRC | 09:05 | |
| *** markstur has joined #openstack-manila | 09:37 | |
| *** markstur has quit IRC | 09:42 | |
| *** markstur has joined #openstack-manila | 10:13 | |
| *** markstur has quit IRC | 10:18 | |
| *** ociuhandu has quit IRC | 10:30 | |
| *** lpetrut has joined #openstack-manila | 10:34 | |
| *** lpetrut_ has joined #openstack-manila | 10:34 | |
| *** markstur has joined #openstack-manila | 10:50 | |
| *** ociuhandu has joined #openstack-manila | 10:51 | |
| *** markstur has quit IRC | 10:55 | |
| *** ganso has joined #openstack-manila | 10:56 | |
| *** markstur has joined #openstack-manila | 11:11 | |
| *** markstur has quit IRC | 11:15 | |
| *** markstur has joined #openstack-manila | 11:48 | |
| *** baojg has joined #openstack-manila | 11:52 | |
| *** markstur has quit IRC | 11:52 | |
| *** MVenesio has joined #openstack-manila | 12:20 | |
| *** MVenesio has quit IRC | 12:20 | |
| *** markstur has joined #openstack-manila | 12:24 | |
| *** markstur has quit IRC | 12:29 | |
| *** gcb has quit IRC | 12:29 | |
| *** gcb has joined #openstack-manila | 12:30 | |
| *** eharney has quit IRC | 12:50 | |
| *** rraja has joined #openstack-manila | 13:00 | |
| *** hoonetorg has quit IRC | 13:08 | |
| *** hoonetorg has joined #openstack-manila | 13:09 | |
| *** catintheroof has joined #openstack-manila | 13:19 | |
| *** dustins has joined #openstack-manila | 13:20 | |
| *** belmoreira has joined #openstack-manila | 13:26 | |
| *** zengyingzhe has quit IRC | 13:26 | |
| *** zengyingzhe has joined #openstack-manila | 13:26 | |
| *** gouthamr has joined #openstack-manila | 13:33 | |
| *** eharney has joined #openstack-manila | 13:41 | |
| *** zengyingzhe_ has joined #openstack-manila | 13:54 | |
| *** zengyingzhe has quit IRC | 13:57 | |
| *** jprovazn has quit IRC | 13:58 | |
| *** ociuhandu has quit IRC | 14:07 | |
| *** baojg has quit IRC | 14:25 | |
| *** baojg has joined #openstack-manila | 14:26 | |
| *** markstur has joined #openstack-manila | 14:35 | |
| *** eharney has quit IRC | 14:35 | |
| *** baojg has quit IRC | 14:37 | |
| *** markstur has quit IRC | 14:39 | |
| *** baojg has joined #openstack-manila | 14:39 | |
| *** baojg has quit IRC | 14:40 | |
| *** baojg has joined #openstack-manila | 14:40 | |
| *** baojg has quit IRC | 14:40 | |
| *** chlong has joined #openstack-manila | 14:41 | |
| *** baojg has joined #openstack-manila | 14:42 | |
| *** baojg has quit IRC | 14:42 | |
| openstackgerrit | Merged openstack/manila master: doc migration: update the doc link address https://review.openstack.org/498359 | 14:43 |
| *** markstur has joined #openstack-manila | 14:44 | |
| *** eharney has joined #openstack-manila | 14:50 | |
| *** jprovazn has joined #openstack-manila | 15:02 | |
| smcginnis | bswartz: Would be good to get your validation on here when you get a chance: https://review.openstack.org/#/c/498263/ | 15:28 |
| bswartz | smcginnis: looking | 15:31 |
| bswartz | smcginnis: never seen diff_start before -- you know what that is? | 15:31 |
| smcginnis | bswartz: That controls the release notes output so you get a full release instead of the separate ones we have now for each RC. | 15:35 |
| bswartz | I see | 15:35 |
| smcginnis | bswartz: There's a small chance of overlap since they do it from the first RC of the last release, but that makes sure all included changes are there. | 15:35 |
| bswartz | so the commit hash in there misses all of the docs changes made in the last week | 15:35 |
| bswartz | I assume that's okay but I wanted to confirm that that was intentional | 15:35 |
| smcginnis | bswartz: Did you do another RC? | 15:35 |
| bswartz | no I was told not to make another RC just for docs changes | 15:36 |
| bswartz | which makes sense because the docs website doesn't care about tags -- just branches | 15:36 |
| smcginnis | Ah, yeah, docs are published separately, so that's ok. As long as there isn't any code change missed that's critical to pike. | 15:36 |
| bswartz | the only other missing changes are the .gitreview file and the upper constrains in tox.ini | 15:37 |
| bswartz | I'm okay to +1 it | 15:38 |
| smcginnis | bswartz: Yep, those are OK too as they aren't included in the release tars. | 15:38 |
| bswartz | done | 15:39 |
| smcginnis | bswartz: Cool, thanks. That will help wrap things up for Pike once we have all the validation from individual projects. | 15:39 |
| bswartz | It will be nice to see pike shipped | 15:39 |
| smcginnis | ++ | 15:40 |
| bswartz | pike was less stressful than some other releases, but still challenging to get done | 15:40 |
| smcginnis | Yeah, same for me. | 15:45 |
| *** xyang_ has joined #openstack-manila | 15:48 | |
| *** rraja has quit IRC | 15:49 | |
| *** lpetrut_ has quit IRC | 15:50 | |
| *** markstur_ has joined #openstack-manila | 15:58 | |
| *** markstur has quit IRC | 16:01 | |
| Reepicheep | is anyone familure with how the manila share service creates openvswitch ports and adds system routes when using the generic backend? | 16:43 |
| Reepicheep | https://bugs.launchpad.net/manila/+bug/1688155 | 16:43 |
| openstack | Launchpad bug 1688155 in Manila "tap device will disappear after manila-share node restart" [Undecided,In progress] - Assigned to yankee (yankeefu) | 16:43 |
| Reepicheep | this seems to be giving me problems ^ ^ | 16:44 |
| Reepicheep | I'm wonder what I need to run to bring these ports / routes backup from the manila share service | 16:45 |
| *** belmoreira has quit IRC | 16:46 | |
| *** belmoreira has joined #openstack-manila | 16:47 | |
| *** belmoreira has quit IRC | 16:48 | |
| *** pcaruana has quit IRC | 16:48 | |
| *** lpetrut has quit IRC | 16:54 | |
| *** raissa has quit IRC | 17:10 | |
| bswartz | Reepicheep: I'm familiar with it | 17:41 |
| bswartz | Reepicheep: you're correct that the generic driver isn't able to recreate things that are destroyed externally -- including ports, tap devices, routes, nova VMs, etc | 17:43 |
| bswartz | work is needed to make the driver able to survive reboots | 17:44 |
| *** lpetrut has joined #openstack-manila | 17:45 | |
| *** catintheroof has quit IRC | 17:49 | |
| *** eharney has quit IRC | 17:49 | |
| *** lpetrut has quit IRC | 17:51 | |
| *** lpetrut has joined #openstack-manila | 17:51 | |
| Reepicheep | bswartz: thanks for the update is there a manual procedure I can use to recreate those ports? | 18:16 |
| Reepicheep | I was looking at the setup_connectivity_with_service_instances function but not being a developer I'm struggling with the steps need | 18:17 |
| *** sumitshatwara has quit IRC | 18:21 | |
| *** catintheroof has joined #openstack-manila | 18:23 | |
| *** xyang_ has quit IRC | 18:34 | |
| *** xyang_ has joined #openstack-manila | 18:35 | |
| *** xyang_ has quit IRC | 18:36 | |
| *** lpetrut has quit IRC | 18:37 | |
| *** eharney has joined #openstack-manila | 18:49 | |
| *** lpetrut has joined #openstack-manila | 18:59 | |
| *** lpetrut has quit IRC | 19:05 | |
| *** catintheroof has quit IRC | 19:30 | |
| *** gcb has quit IRC | 19:33 | |
| *** lpetrut has joined #openstack-manila | 19:33 | |
| *** catintheroof has joined #openstack-manila | 19:36 | |
| *** jprovazn has quit IRC | 19:43 | |
| *** xyang1 has joined #openstack-manila | 19:47 | |
| *** belmoreira has joined #openstack-manila | 20:04 | |
| bswartz | Reepicheep: I've never tried it -- basically you'd have to follow the code and replicate what it does | 20:44 |
| bswartz | we maybe need to clarify the existing limitations of the generic driver in our docs | 20:45 |
| bswartz | we've designed the generic driver to be production ready eventually, but there are gaps to be addressed before we get there | 20:46 |
| *** ChanServ changes topic to "OpenStack Shared File Systems | Manila | PTG Sept 20-21 https://etherpad.openstack.org/p/manila-ptg-queens" | 20:47 | |
| bswartz | Reepicheep: I put a topic on the PTG etherpad to discuss it in September | 20:48 |
| *** gouthamr has quit IRC | 21:05 | |
| Reepicheep | bswartz: thanks.. I'm going to have to try to figure out what it does. I'm basically in a spot that I need to get it back online | 21:24 |
| Reepicheep | generic is currently one of the few options that I have :( | 21:28 |
| Reepicheep | am I correct in understanding that the cephfs driver still has some significant work to be done also? | 21:30 |
| *** catintheroof has quit IRC | 21:34 | |
| *** tbarron is now known as tbarron|errand | 21:46 | |
| *** belmoreira has quit IRC | 21:49 | |
| *** markstur_ has quit IRC | 21:50 | |
| *** lpetrut has quit IRC | 21:55 | |
| *** xyang1 has quit IRC | 22:06 | |
| *** gouthamr has joined #openstack-manila | 22:38 | |
| *** dustins has quit IRC | 22:45 | |
| *** ianychoi has quit IRC | 23:17 | |
| *** ianychoi has joined #openstack-manila | 23:17 | |
| *** jmlowe has joined #openstack-manila | 23:43 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!