| *** david-lyle has quit IRC | 00:05 | |
| openstackgerrit | Merged openstack/manila: [Unity driver] VLAN enhancement https://review.openstack.org/410037 | 00:07 |
|---|---|---|
| *** catintheroof has joined #openstack-manila | 00:23 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/python-manilaclient: [DNM] Testing this against the access rule API refactor https://review.openstack.org/422307 | 00:24 |
| *** kaisers has joined #openstack-manila | 00:27 | |
| *** harlowja has joined #openstack-manila | 00:52 | |
| *** kaisers has quit IRC | 00:54 | |
| *** catintheroof has quit IRC | 01:13 | |
| *** catintheroof has joined #openstack-manila | 01:14 | |
| *** catintheroof has quit IRC | 01:14 | |
| *** tuanluong has joined #openstack-manila | 01:23 | |
| *** mtanino has quit IRC | 01:24 | |
| *** cknight has quit IRC | 01:26 | |
| *** lgreg1 has quit IRC | 01:41 | |
| *** kaisers has joined #openstack-manila | 01:50 | |
| *** kaisers_ has joined #openstack-manila | 02:10 | |
| *** kaisers has quit IRC | 02:13 | |
| *** gcb has joined #openstack-manila | 02:14 | |
| *** kaisers_ has quit IRC | 02:18 | |
| *** gouthamr_ has joined #openstack-manila | 02:35 | |
| *** darrenc_ has joined #openstack-manila | 02:35 | |
| *** sage_ has joined #openstack-manila | 02:37 | |
| *** gouthamr has quit IRC | 02:41 | |
| *** sage has quit IRC | 02:41 | |
| *** vkmc has quit IRC | 02:41 | |
| *** darrenc has quit IRC | 02:41 | |
| *** tbarron has quit IRC | 02:41 | |
| *** gouthamr_ is now known as gouthamr | 02:44 | |
| *** tbarron has joined #openstack-manila | 03:02 | |
| *** vkmc has joined #openstack-manila | 03:06 | |
| *** lgreg has joined #openstack-manila | 03:57 | |
| *** cknight has joined #openstack-manila | 04:00 | |
| *** kaisers has joined #openstack-manila | 04:14 | |
| *** absubram has joined #openstack-manila | 04:33 | |
| *** kaisers has quit IRC | 04:42 | |
| *** furlongm_ has quit IRC | 04:44 | |
| *** furlongm_ has joined #openstack-manila | 04:44 | |
| *** absubram has quit IRC | 05:03 | |
| *** cknight has quit IRC | 05:03 | |
| *** kaisers has joined #openstack-manila | 05:39 | |
| *** kaisers has quit IRC | 05:52 | |
| *** kaisers has joined #openstack-manila | 05:52 | |
| *** kaisers has joined #openstack-manila | 05:53 | |
| *** kaisers has quit IRC | 05:58 | |
| *** lgreg has quit IRC | 06:02 | |
| *** gouthamr has quit IRC | 06:04 | |
| *** kaisers has joined #openstack-manila | 06:23 | |
| *** kaisers has quit IRC | 06:28 | |
| *** lpetrut has joined #openstack-manila | 07:08 | |
| *** jprovazn has joined #openstack-manila | 07:17 | |
| *** kaisers has joined #openstack-manila | 07:24 | |
| *** kaisers has quit IRC | 07:26 | |
| *** kaisers has joined #openstack-manila | 07:26 | |
| *** lpetrut has quit IRC | 07:54 | |
| *** lpetrut has joined #openstack-manila | 08:04 | |
| *** ociuhandu has quit IRC | 08:18 | |
| *** lpetrut has quit IRC | 08:26 | |
| *** sapcc-bot has quit IRC | 08:30 | |
| *** carthaca_ has joined #openstack-manila | 08:30 | |
| *** sapcc-bot2 has joined #openstack-manila | 08:30 | |
| *** dgonzalez_ has joined #openstack-manila | 08:30 | |
| *** tpatzig_ has joined #openstack-manila | 08:30 | |
| *** mkoderer_ has joined #openstack-manila | 08:30 | |
| *** mkoderer_ has quit IRC | 08:32 | |
| *** dgonzalez_ has quit IRC | 08:32 | |
| *** carthaca_ has quit IRC | 08:32 | |
| *** tpatzig_ has quit IRC | 08:32 | |
| *** openstackgerrit has quit IRC | 08:33 | |
| *** rraja has joined #openstack-manila | 08:35 | |
| *** sapcc-bot has joined #openstack-manila | 08:42 | |
| *** sapcc-bot2 has quit IRC | 08:42 | |
| *** mkoderer_ has joined #openstack-manila | 08:42 | |
| *** dgonzalez_ has joined #openstack-manila | 08:42 | |
| *** tpatzig_ has joined #openstack-manila | 08:42 | |
| *** carthaca_ has joined #openstack-manila | 08:42 | |
| *** mkoderer_ has quit IRC | 08:44 | |
| *** dgonzalez_ has quit IRC | 08:44 | |
| *** tpatzig_ has quit IRC | 08:44 | |
| *** carthaca_ has quit IRC | 08:44 | |
| *** kaisers has quit IRC | 08:46 | |
| *** kaisers has joined #openstack-manila | 08:46 | |
| *** jprovazn has quit IRC | 08:52 | |
| *** jprovazn has joined #openstack-manila | 09:03 | |
| *** ociuhandu has joined #openstack-manila | 09:08 | |
| *** zhonghua has quit IRC | 09:40 | |
| *** zhonghua has joined #openstack-manila | 09:40 | |
| *** a-pugachev has joined #openstack-manila | 09:52 | |
| *** tuanluong has quit IRC | 09:56 | |
| *** yumiriam has joined #openstack-manila | 10:01 | |
| *** lpetrut has joined #openstack-manila | 10:03 | |
| *** ganso has joined #openstack-manila | 10:04 | |
| ganso | bswartz: pong | 10:05 |
| *** lpetrut has quit IRC | 10:55 | |
| *** lpetrut has joined #openstack-manila | 10:55 | |
| *** sapcc-bot has quit IRC | 11:00 | |
| *** sapcc-bot has joined #openstack-manila | 11:00 | |
| *** carthaca_ has joined #openstack-manila | 11:00 | |
| *** dgonzalez_ has joined #openstack-manila | 11:00 | |
| *** tpatzig_ has joined #openstack-manila | 11:00 | |
| *** openstackgerrit has joined #openstack-manila | 11:01 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/python-manilaclient: Add share group support to Manila client https://review.openstack.org/335120 | 11:01 |
| *** carthaca_ has quit IRC | 11:02 | |
| *** dgonzalez_ has quit IRC | 11:02 | |
| *** tpatzig_ has quit IRC | 11:02 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Rename consistency group modules to share groups https://review.openstack.org/409864 | 11:02 |
| *** zhugaoxiao has joined #openstack-manila | 11:14 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Manila Share Groups https://review.openstack.org/335093 | 11:15 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: [Tempest] Add functional tests for share groups feature https://review.openstack.org/355264 | 11:15 |
| *** alyson_ has joined #openstack-manila | 11:21 | |
| openstackgerrit | Alyson proposed openstack/python-manilaclient: Add mountable snapshots support to manila client https://review.openstack.org/345625 | 11:24 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support https://review.openstack.org/345526 | 11:27 |
| *** zhugaoxiao has quit IRC | 11:29 | |
| *** zhugaoxiao has joined #openstack-manila | 11:29 | |
| ganso | vponomaryov, tbarron, markstur: Hello. Could you please review this driver implementation of revert-to-snapshot when you have some review time? https://review.openstack.org/#/c/411923/ Thanks in advance! | 11:37 |
| openstackgerrit | yankee proposed openstack/python-manilaclient: Support instance's status of migrating in client https://review.openstack.org/414829 | 11:45 |
| *** JoseMello has joined #openstack-manila | 12:01 | |
| vponomaryov | ganso: have you considered "reusing of 'share' tables related to export locations" implementing "mountable snapshots" feature? | 12:10 |
| ganso | vponomaryov: I haven't implemented it, tpsilva did | 12:10 |
| ganso | vponomaryov: so I don't know if he considered | 12:11 |
| vponomaryov | ganso: ok, what do you think about it? | 12:11 |
| ganso | vponomaryov: I think the main problem would be the mappings | 12:11 |
| ganso | vponomaryov: it is currently mapping to share instances | 12:12 |
| vponomaryov | also, existing spec https://github.com/openstack/manila-specs/blob/master/specs/ocata/mountable-snapshots.rst omits too many valueable data | 12:12 |
| vponomaryov | related to API and DB structure | 12:13 |
| ganso | vponomaryov: probably data that was found out later during implementation, but what's in the spec has been agreed on | 12:13 |
| vponomaryov | ganso: there is no APi responses | 12:14 |
| vponomaryov | ganso: there is no DB table structure | 12:14 |
| ganso | vponomaryov: yes, I see it | 12:14 |
| vponomaryov | ganso: so, I think we do not have "real" agreement here | 12:14 |
| vponomaryov | at all | 12:14 |
| ganso | vponomaryov: indeed, I do not see your +2 on the spec patch | 12:15 |
| ganso | vponomaryov: oh wait, I do | 12:16 |
| ganso | vponomaryov: was looking at wrong patch | 12:16 |
| ganso | vponomaryov: so I guess there was agreement after all | 12:16 |
| vponomaryov | ganso: i am saying no one really understands it | 12:16 |
| vponomaryov | ganso: valueable parts are just absent | 12:17 |
| vponomaryov | ganso: possible approach to reuse 'share' tables is not considered | 12:17 |
| ganso | vponomaryov: I can make a spec update, similar to what I am doing for migration, but only after the main patch merges | 12:17 |
| vponomaryov | ganso: and not described why it should not be used | 12:17 |
| vponomaryov | ganso: only after? )) | 12:17 |
| vponomaryov | ^_^ | 12:17 |
| ganso | vponomaryov: we can decide all this by reviewing the main patch | 12:17 |
| vponomaryov | yes, we can, it is already going this way | 12:18 |
| ganso | vponomaryov: at this point, there is no need to debate on the spec, it has been agreed in the past, and possible changes when implementing it are expected | 12:18 |
| vponomaryov | BUT | 12:18 |
| vponomaryov | time | 12:18 |
| *** kaisers has quit IRC | 12:20 | |
| *** kaisers has joined #openstack-manila | 12:21 | |
| *** kaisers has quit IRC | 12:26 | |
| *** catintheroof has joined #openstack-manila | 12:43 | |
| ganso | bswartz: ping | 12:49 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support to HNAS driver https://review.openstack.org/411474 | 12:51 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Mountable snapshots scenario tests https://review.openstack.org/412001 | 12:51 |
| *** gcb has quit IRC | 12:59 | |
| *** jcsp has joined #openstack-manila | 13:08 | |
| *** gouthamr has joined #openstack-manila | 13:12 | |
| gouthamr | ganso: ping | 13:15 |
| gouthamr | vponomaryov: ping | 13:23 |
| *** kaisers has joined #openstack-manila | 13:27 | |
| *** kaisers has quit IRC | 13:28 | |
| *** kaisers has joined #openstack-manila | 13:29 | |
| *** kaisers2 has quit IRC | 13:31 | |
| *** kaisers3 has quit IRC | 13:31 | |
| *** sandanar has joined #openstack-manila | 13:42 | |
| *** tommylikehu_ has joined #openstack-manila | 13:43 | |
| *** kaisers1 has joined #openstack-manila | 13:43 | |
| *** tommylikehu_ has quit IRC | 13:44 | |
| *** tommylikehu_ has joined #openstack-manila | 13:44 | |
| *** tommylikehu_ has quit IRC | 13:45 | |
| *** kaisers2 has joined #openstack-manila | 13:45 | |
| *** tommylikehu_ has joined #openstack-manila | 13:46 | |
| *** eharney has joined #openstack-manila | 13:49 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Refactor Access Rules APIs https://review.openstack.org/369668 | 13:58 |
| vponomaryov | gouthamr: pong | 14:04 |
| *** porrua has joined #openstack-manila | 14:05 | |
| *** cknight has joined #openstack-manila | 14:07 | |
| *** porrua has quit IRC | 14:09 | |
| gouthamr | vponomaryov: hey! question regarding a bug you raised a long time ago: https://bugs.launchpad.net/manila/+bug/1512403 | 14:11 |
| openstack | Launchpad bug 1512403 in Manila "API actions can not be versioned" [Low,Confirmed] | 14:11 |
| vponomaryov | gouthamr: yes? | 14:12 |
| gouthamr | vponomaryov: was wondering what really fails here... i tried using the version decorator to the allow_access methods and running some tests, i get the behavior expected.. | 14:12 |
| vponomaryov | gouthamr: tommylikehu says the same | 14:13 |
| gouthamr | vponomaryov: yes, i'm following up after his comment.. | 14:13 |
| vponomaryov | gouthamr: looks like it was fixed without regards to this report | 14:13 |
| vponomaryov | gouthamr: but I am not aware about it | 14:13 |
| vponomaryov | try manila as of that state | 14:14 |
| vponomaryov | 2015-11-02 | 14:14 |
| vponomaryov | if oyu want to see bug | 14:14 |
| *** mkoderer has joined #openstack-manila | 14:14 | |
| gouthamr | vponomaryov: i don't see a fix committed.. or any changes that stand out in api/openstack/wsgi.py | 14:15 |
| *** kaisers has quit IRC | 14:15 | |
| *** kaisers has joined #openstack-manila | 14:16 | |
| gouthamr | vponomaryov: hmmm, okay will do.. | 14:16 |
| gouthamr | vponomaryov: thanks! | 14:16 |
| gouthamr | vponomaryov: though i don't think we want to use the decorator if the code is more readable without it | 14:19 |
| vponomaryov | gouthamr: you mean usage of conditions with microversion comparison? | 14:19 |
| vponomaryov | gouthamr: if yes, then you know why it had "low" priority | 14:20 |
| gouthamr | vponomaryov: yeah makes sense.. i'll confirm that it was fixed as intended and mark that bug appropriately.. | 14:21 |
| vponomaryov | gouthamr: also, keeping in mind that it is low priority, I would suggest not spending time on it prior to FF | 14:21 |
| gouthamr | vponomaryov: +1 | 14:22 |
| ganso | gouthamr: pong | 14:27 |
| gouthamr | ganso: ^ same question :) | 14:27 |
| ganso | gouthamr: ok | 14:28 |
| ganso | gouthamr: can you weigh in in the migration patch? regarding hyphens vs a underscore | 14:28 |
| ganso | gouthamr: I posted a response there | 14:28 |
| gouthamr | ganso: i did...? | 14:28 |
| bswartz | ganso: good morning | 14:29 |
| bswartz | ganso: I couldn't find you yesterday afternoon | 14:29 |
| gouthamr | ganso: yeah, while i like bswartz's suggestion, we can't break consistency.. all of our APIs use _s in parameters. | 14:29 |
| ganso | bswartz: morning | 14:29 |
| ganso | bswartz: I had left the office | 14:29 |
| ganso | gouthamr: we can't break or we should break? | 14:30 |
| ganso | bswartz: I posted several responses | 14:30 |
| ganso | bswartz: guess we have a discussion ahead of us | 14:30 |
| bswartz | gouthamr: consistency with what? | 14:31 |
| gouthamr | bswartz: consistency with all of manila APIs that send any POST data | 14:32 |
| bswartz | gouthamr: what about other openstack services? is there an official style guide for this kind of thing? | 14:32 |
| gouthamr | bswartz: no i tried to look up one.. maybe we should propose it, afaics, we followed nova closely and nova does the same for API params: http://developer.openstack.org/api-ref/compute/?expanded=create-server-detail | 14:33 |
| gouthamr | bswartz: https://specs.openstack.org/openstack/api-wg/guidelines/naming.html#fields-in-an-api-request-or-response-body | 14:34 |
| gouthamr | bswartz: i referred to this in my email response: "Field names should use the snake_case style, not CamelCase or StUdLyCaPs style." | 14:35 |
| bswartz | yeah I read it | 14:35 |
| bswartz | well if snake case is preferred I can withdraw my objection | 14:35 |
| bswartz | I just think that snake case is horrifying in HTTP bodies | 14:35 |
| bswartz | I guess my opinion is not shared | 14:36 |
| vponomaryov | gouthamr: export locations for shares violate rules | 14:37 |
| vponomaryov | gouthamr: they use underscore as part of the url | 14:37 |
| gouthamr | vponomaryov: yes, which was the point of your email iirc, we didn't decide if we'd resolve that.. | 14:37 |
| gouthamr | vponomaryov: we probably should leave it alone... deprecating an API because we used "_" in the resource name is probably overkill | 14:38 |
| vponomaryov | gouthamr: we should add alias | 14:39 |
| vponomaryov | gouthamr: proper alias | 14:39 |
| vponomaryov | gouthamr: and use it in client | 14:39 |
| vponomaryov | and tempest | 14:39 |
| gouthamr | vponomaryov: +1 | 14:39 |
| *** Yogi1 has joined #openstack-manila | 14:41 | |
| vponomaryov | tbarron: we already have a lot of big methods | 14:42 |
| tbarron | vponomaryov: so? | 14:42 |
| tbarron | vponomaryov: if you can understand it and are confident it works +2 | 14:42 |
| vponomaryov | tbarron: yes, my brain hurts when I see this change )) | 14:43 |
| vponomaryov | tbarron: but we have tests | 14:43 |
| vponomaryov | tbarron: functional tests | 14:43 |
| tbarron | tests are not a substitute for being able to understand the code | 14:43 |
| vponomaryov | tbarron: to make sure it works | 14:43 |
| vponomaryov | tbarron: agree | 14:43 |
| vponomaryov | tbarron: in good case, whole this change should be split up to lots of smaller changes | 14:44 |
| tbarron | and they probably won't catch the races that we are fixing here | 14:44 |
| *** dustins has joined #openstack-manila | 14:44 | |
| tbarron | vponomaryov: agree, but I'm not insisting on that | 14:44 |
| tbarron | in general the code in this patch is quite clean | 14:44 |
| vponomaryov | tbarron: clean, just mixing over9000 logical changes ) | 14:44 |
| tbarron | but in probably the most critical part we've made a big ugly hard to understand method even bigger and harder to understand | 14:45 |
| tbarron | I think the general idea is good, I just can't verify that its implementation is correct. | 14:45 |
| vponomaryov | tbarron: that is why you should insist on splitting this up ) | 14:46 |
| vponomaryov | tbarron: if you cannot see the whole picture | 14:46 |
| vponomaryov | tbarron: but I guess gouthamr will be upset ) | 14:46 |
| tbarron | I am insisting that we split up this method. | 14:46 |
| tbarron | vponomaryov: as you know I don't worry too much about upsetting people. | 14:47 |
| gouthamr | was upset writing that method, trust me | 14:47 |
| gouthamr | lol | 14:47 |
| tbarron | I have customers | 14:47 |
| tbarron | They being upsset is more my concern. | 14:47 |
| tommylikehu_ | ping zengyingzhe | 14:48 |
| vponomaryov | gouthamr: mission impossible - split it up so everyone is happy to merge it right away =) | 14:48 |
| ganso | vponomaryov: Could you please elaborate on why you are requesting this? https://review.openstack.org/#/c/345526/44/manila/api/v2/share_snapshots.py | 14:48 |
| ganso | vponomaryov: share export locations is separate from share module | 14:49 |
| gouthamr | vponomaryov: yessir. onnit. :) it's pretty easy to split up.. maybe i just got used to it with all these patchsets.. too bad i was the only one getting used to it :P | 14:49 |
| ganso | gouthamr: I got used to it too... | 14:49 |
| vponomaryov | ganso: do you have the answer for question raised there? | 14:49 |
| gouthamr | ganso: :D i know you did. gave you some inspiration for all that nice beer | 14:50 |
| *** breitz has quit IRC | 14:50 | |
| ganso | vponomaryov: answer is because it looks good, as share module | 14:50 |
| *** breitz has joined #openstack-manila | 14:50 | |
| ganso | gouthamr: lol xD | 14:50 |
| vponomaryov | ganso: not strong enough argument | 14:50 |
| ganso | vponomaryov: also, consistency | 14:50 |
| vponomaryov | ganso: better to rewrite share's part too | 14:51 |
| ganso | vponomaryov: what's the problem with it? | 14:51 |
| ganso | vponomaryov: it is better organized this way | 14:51 |
| ganso | vponomaryov: and access rules do not have its own controllers, you don't have "get" access rule API actions, just index | 14:51 |
| *** dustins has quit IRC | 14:51 | |
| *** dustins has joined #openstack-manila | 14:52 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Refactor Access Rules APIs https://review.openstack.org/369668 | 14:55 |
| vponomaryov | ganso: "access rules" are made in the same way as export locations - nested objects | 14:56 |
| vponomaryov | ganso: so, export locations could be implemented using completely the same approach | 14:57 |
| ganso | vponomaryov: could, but if it is more organized implemented this way, why shove everything under a single module? | 14:58 |
| *** mtanino has joined #openstack-manila | 14:58 | |
| vponomaryov | ganso: you handle alike-objects in different ways | 14:59 |
| vponomaryov | ganso: it cannot be "more organized" | 14:59 |
| vponomaryov | ganso: it is exactly vie versa | 14:59 |
| vponomaryov | s/vie versa/vice versa/ | 14:59 |
| ganso | vponomaryov: how? | 15:00 |
| ganso | vponomaryov: what would be the advantages of merging the share_snapshot_export_locations and instance modules with the share_snapshot module? | 15:01 |
| vponomaryov | later | 15:01 |
| *** xyang_ has joined #openstack-manila | 15:02 | |
| *** a-pugachev has quit IRC | 15:05 | |
| *** a-pugachev has joined #openstack-manila | 15:06 | |
| *** a-pugachev has quit IRC | 15:07 | |
| *** rraja has quit IRC | 15:18 | |
| *** makowals has joined #openstack-manila | 15:21 | |
| *** porrua has joined #openstack-manila | 15:21 | |
| *** porrua has quit IRC | 15:26 | |
| *** lgreg has joined #openstack-manila | 15:29 | |
| *** timcl has quit IRC | 15:45 | |
| *** mtanino has quit IRC | 15:46 | |
| *** lpetrut has quit IRC | 15:50 | |
| *** tommylikehu_ has quit IRC | 15:54 | |
| *** timcl has joined #openstack-manila | 15:59 | |
| gouthamr | ganso: "is there any way to have more than one snapshot instance on a non-replicated share?" -> no | 16:01 |
| ganso | gouthamr: great | 16:01 |
| gouthamr | ganso: i'm assuming you meant one snapshot instance per snapshot | 16:01 |
| ganso | gouthamr: guess I coded that snapshot mapping thing in advance xD | 16:01 |
| bswartz | ganso: I don't see where we're disagreeing | 16:02 |
| bswartz | errors occur rarely, but when they do, they're fatal | 16:02 |
| ganso | bswartz: so, right now, for migration, your comment is not valid... but you bring up a valid point where the status of replicated snapshots are not consistent with share replicas | 16:02 |
| bswartz | hold on I need to go back and look at my comment | 16:03 |
| gouthamr | ganso: how? we always pick up the active share instance's snapshot instance as the snapshot | 16:03 |
| ganso | gouthamr: that's why, the order = (constants.STATUS_MIGRATING, constants.STATUS_AVAILABLE, constants.STATUS_ERROR) does not matter because we only have one | 16:04 |
| bswartz | ganso: we will we have during an active migration or after a failed migration? | 16:05 |
| ganso | bswartz: that order of status we are talking about is for the active instance, the one that is shown when user does snapshot-list while a migration is in progress | 16:05 |
| ganso | bswartz: we should always show MIGRATING in this case, that's why it is on top | 16:06 |
| bswartz | okay my comment was on the model itself -- so it will affect all snapshots that have 2 or more instances | 16:06 |
| *** david-lyle has joined #openstack-manila | 16:06 | |
| ganso | bswartz: regarding AVAILABLE vs ERROR, it will relate only to replicated snapshots, which are not applicable to migration at all | 16:06 |
| bswartz | ganso: you comment was "does not matter because we only have one" -- won't we have 2 while migrating? | 16:07 |
| ganso | bswartz: yes, but we should always show MIGRATING | 16:07 |
| bswartz | migration should sort higher than available, so it should should migrating normally | 16:07 |
| gouthamr | bswartz: we will have two when we're migrating, but we pick up the one that's 'migrating' | 16:07 |
| bswartz | however once an error occurs, should migrating sort higher than error? | 16:07 |
| bswartz | because an error should stop the migration IMO | 16:08 |
| ganso | bswartz: still should, IMO | 16:08 |
| ganso | bswartz: it does | 16:08 |
| bswartz | so what happens? | 16:08 |
| gouthamr | bswartz: once an error occurs where/ | 16:08 |
| ganso | bswartz: an exception during migration will be raised | 16:09 |
| ganso | bswartz: this error could happen in migration-start/continue or migration-complete | 16:09 |
| ganso | bswartz: if it happens in start/continue, it is cleaned up, the source snapshot instance is revert back to AVAILABLE, and the destination is deleted (we don't care about the destination in a failed migration that did not reach migration-complete, it can be discarded as everything is preserved in the source) | 16:10 |
| bswartz | okay that's importnat | 16:10 |
| bswartz | is there any case where it doesn't get cleaned up automatically? | 16:10 |
| ganso | bswartz: yes, if migration-complete fails during a driver-assisted migration | 16:11 |
| *** carthaca_ has joined #openstack-manila | 16:11 | |
| *** tpatzig_ has joined #openstack-manila | 16:11 | |
| bswartz | in that case you're left with 2 instances? | 16:11 |
| bswartz | and what would their states be? | 16:11 |
| ganso | bswartz: no, the source is set back to available, the destination is set to error, but not deleted | 16:12 |
| ganso | bswartz: sorry, yes, 2 instances | 16:12 |
| ganso | bswartz: the source instance is revert back to available so it can be used | 16:12 |
| ganso | bswartz: if stuck in migrating status, it is unusable for the user | 16:12 |
| gouthamr | ganso: if the source share is set to migration_error, why can't you set the source snapshots to the same? why 'available'? | 16:13 |
| *** tpatzig_ has quit IRC | 16:13 | |
| *** carthaca_ has quit IRC | 16:13 | |
| gouthamr | ganso: i.e, how can you guarantee that migration didn't mess anything up? | 16:13 |
| ganso | gouthamr: we can't, that's why we don't delete | 16:13 |
| ganso | gouthamr: the source share is set to 'available', not 'error' gouthamr | 16:13 |
| bswartz | but we want to prevent the user from doing anything to the share until it's cleaned up | 16:14 |
| gouthamr | ganso: ? we do | 16:14 |
| ganso | gouthamr: / | 16:14 |
| ganso | gouthamr: ? | 16:14 |
| gouthamr | ganso: sorry i missed that.. we have a migration_error status for the same reason | 16:14 |
| *** zhonghua2 has joined #openstack-manila | 16:14 | |
| ganso | gouthamr: we have the task state | 16:14 |
| ganso | gouthamr: not status | 16:14 |
| gouthamr | ganso: if the share is available, it can be used by others.. | 16:14 |
| bswartz | okay so the problem here is that we're discussion code that affects all snapshots in the context of one very specific workflow | 16:15 |
| ganso | bswartz: yes | 16:15 |
| bswartz | I think I'm convinced by ganso that nothing bad can happen here with migrations of snapshots | 16:15 |
| bswartz | but the next feature to use snapshot instances might be different | 16:15 |
| bswartz | and this code is in the code snapshot model | 16:15 |
| *** zhonghua has quit IRC | 16:16 | |
| ganso | bswartz: so, there was no order at all before | 16:17 |
| ganso | bswartz: an order had to be added for MIGRATING to be on top | 16:17 |
| ganso | bswartz: thing is, the current order for shares is available on top of error | 16:17 |
| bswartz | and if we follow my suggestion then what bad thing happens? | 16:17 |
| ganso | bswartz: the user will see the snapshot in error state, while the share is in available state, which is inconsistent | 16:18 |
| gouthamr | ganso: have you considered the aggregate status below? | 16:18 |
| ganso | bswartz: if we are changing the order, the share instances order has to be changed as well | 16:18 |
| ganso | gouthamr: that is one confusing field | 16:19 |
| bswartz | ganso: not at all | 16:19 |
| bswartz | you can have error snapshots on normal shares | 16:19 |
| bswartz | any time you fail to create a snapshot it will be in error state | 16:19 |
| ganso | gouthamr: I am not sure I understand the difference between aggregate status and the regular snapshot['status'] which already aggregates snapshot instances status (it shouldn't do snapshot.instances[0] at all) | 16:19 |
| gouthamr | ganso: aggregate status is what is used in the view builder | 16:20 |
| gouthamr | ganso: taht's what the user sees | 16:20 |
| ganso | gouthamr: why do we have this? | 16:20 |
| ganso | gouthamr: and not snapshot.instance that selects the proper instance? | 16:20 |
| bswartz | because snapshots don't have statuses, snapshots instances do | 16:21 |
| ganso | bswartz: same as shares' | 16:21 |
| *** sandanar has quit IRC | 16:21 | |
| ganso | bswartz: we do not have aggregate statuses on share's, we select the proper instance when one does share.instance | 16:21 |
| bswartz | errr | 16:22 |
| *** a-pugachev has joined #openstack-manila | 16:22 | |
| bswartz | okay so maybe that needs fixing | 16:22 |
| bswartz | they should work the same | 16:22 |
| *** absubram has joined #openstack-manila | 16:22 | |
| bswartz | I was only reviewing the new code in ganso's patch | 16:22 |
| ganso | bswartz: yes, I was trying to get them to work the same | 16:23 |
| ganso | bswartz: I also found it every weird that the aggregate status in snapshots are in opposed order as the share's? | 16:23 |
| ganso | bswartz: s/?/ | 16:23 |
| * bswartz sighs | 16:24 | |
| ganso | bswartz: s/every/very | 16:24 |
| bswartz | looks like for shares the error instances are sorted last? | 16:24 |
| ganso | bswartz: exactly | 16:24 |
| * bswartz headdesk | 16:24 | |
| ganso | lol | 16:24 |
| gouthamr | misses u_glide | 16:24 |
| bswartz | I need my git blame | 16:24 |
| ganso | so, I remember gouthamr and I touched that order last | 16:25 |
| ganso | that's why it is important that gouthamr is here so we can undestand why it is like this | 16:25 |
| gouthamr | as was necessary with migration and replication, only uses of share instances so far | 16:25 |
| bswartz | cknight's name is on that order list | 16:26 |
| bswartz | gouthamr's on the rest of that function | 16:26 |
| ganso | gouthamr: yes, so if we agreed to that in the past, why would snapshot's order be any different? | 16:26 |
| bswartz | Let me see if I can find out why we made this decision for share instances | 16:27 |
| gouthamr | ganso: we use snapshot statuses to determine the health across replicas | 16:27 |
| gouthamr | the user doesn't have APIs to check the exact snapshot replicas | 16:27 |
| bswartz | it's lunchtime here and I'm not able to argue on an empty stomach | 16:27 |
| ganso | gouthamr: what about the health of a share across replicas? | 16:27 |
| gouthamr | there are share replica APIs | 16:27 |
| ganso | so I guess we are missing an API then | 16:28 |
| ganso | IMO this field is being overloaded and due to that, possibly misused | 16:28 |
| gouthamr | yes, aggregate_status could be bandaid code for some late-breaking design decisions in mitaka.. however, if we haven't changed it since, i don't see how it's broken.. | 16:28 |
| * ganso is referring to the 'aggregate_status' | 16:28 | |
| gouthamr | it's not | 16:29 |
| gouthamr | it was built specifically for the view builder on snapshots | 16:29 |
| gouthamr | user lists a snapshot of a share - needs to know if it's truly "available" across all instances of the share | 16:29 |
| ganso | gouthamr: why would the snapshot viewbuilder result be different from snapshot.instance? | 16:30 |
| ganso | gouthamr: 'status': snapshot.get('aggregate_status'), | 16:31 |
| ganso | gouthamr: if it was a separate field that would be awesome and not confusing at all | 16:32 |
| gouthamr | i take back what i said about the view builder | 16:32 |
| gouthamr | it's used as the snapshot's status everywhere else (and that is correct) | 16:32 |
| gouthamr | though, no, i don't have an answer right off the top of my head why this can't be instance that we pick for a snapshot | 16:33 |
| ganso | gouthamr: except snapshot.instance, which is mostly used internally for all actions | 16:33 |
| ganso | gouthamr: ok, I think we need a separate patch to fix that | 16:33 |
| ganso | gouthamr: I guess the code to select the instance in my patch is correct, because snapshot.instance is used for actions. But maybe aggregate status should have MIGRATING on top as well then | 16:35 |
| ganso | gouthamr: it is not the solution to the problem, but it also shouldn't be fixed within my patch, I only need to make the changes for migration to work | 16:35 |
| gouthamr | ganso: hmmm... i think i see an answer here: aggregate status is used for status checks all over.. and is most important in the fragile exchange between the driver and the share manager in replication operations | 16:40 |
| ganso | gouthamr: so, it must not be touched? how can MIGRATING fit in there? | 16:41 |
| gouthamr | ganso: you'd add it before 'available' | 16:42 |
| ganso | gouthamr: ok, thanks | 16:43 |
| gouthamr | ganso: and make sure to create the destination snapshot instance with 'migrating' status as well.. | 16:43 |
| ganso | gouthamr: it is MIGRATING_TO | 16:43 |
| gouthamr | ganso: yes.. | 16:43 |
| *** absubram has quit IRC | 16:47 | |
| ganso | gouthamr: are there more updates incoming to the refactor access rules patch? | 16:48 |
| gouthamr | ganso: no | 16:48 |
| ganso | gouthamr: ok, thanks | 16:48 |
| *** absubram has joined #openstack-manila | 16:49 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Tooz integration https://review.openstack.org/318336 | 16:51 |
| *** david-lyle has quit IRC | 16:55 | |
| *** absubram has quit IRC | 16:55 | |
| *** absubram has joined #openstack-manila | 16:59 | |
| *** timcl has quit IRC | 17:00 | |
| *** xyang_ has quit IRC | 17:01 | |
| *** absubram has quit IRC | 17:08 | |
| *** timcl has joined #openstack-manila | 17:08 | |
| openstackgerrit | Vitaliy Levitski proposed openstack/manila: debug https://review.openstack.org/422769 | 17:10 |
| *** tommylikehu1 has joined #openstack-manila | 17:15 | |
| *** zengyingzhe_ has joined #openstack-manila | 17:15 | |
| *** tommylikehu has quit IRC | 17:18 | |
| *** tommylikehu1 is now known as tommylikehu | 17:18 | |
| *** david-lyle has joined #openstack-manila | 17:18 | |
| *** david-lyle has quit IRC | 17:18 | |
| *** zengyingzhe has quit IRC | 17:18 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support https://review.openstack.org/345526 | 17:19 |
| *** kaisers has quit IRC | 17:21 | |
| *** lpetrut has joined #openstack-manila | 17:27 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support https://review.openstack.org/345526 | 17:27 |
| *** david-lyle has joined #openstack-manila | 17:28 | |
| *** mkoderer has quit IRC | 17:36 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support https://review.openstack.org/345526 | 17:37 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support https://review.openstack.org/345526 | 17:55 |
| *** absubram has joined #openstack-manila | 17:56 | |
| *** xyang_ has joined #openstack-manila | 17:58 | |
| *** lgreg has left #openstack-manila | 18:11 | |
| *** david-lyle has quit IRC | 18:18 | |
| *** catinthe_ has joined #openstack-manila | 18:18 | |
| *** catintheroof has quit IRC | 18:18 | |
| *** ociuhandu has quit IRC | 18:35 | |
| *** absubram has quit IRC | 18:36 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Share Migration Ocata Improvements https://review.openstack.org/406305 | 18:37 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/python-manilaclient: Implement Share Migration Ocata improvements https://review.openstack.org/406306 | 18:42 |
| bswartz | ganso: So I dug into the history of the share instance / snapshot instance evolution | 18:49 |
| bswartz | ganso: gouthamr pointed out a few things that are important I Think | 18:49 |
| bswartz | ganso: would you rather discuss here or in a meeting of some kind? | 18:49 |
| ganso | bswartz: we can discuss here | 18:50 |
| bswartz | k | 18:50 |
| *** a-pugachev has quit IRC | 18:50 | |
| bswartz | so our first observation is that the sort order for share instances is probably wrong, and error should come first in that list | 18:51 |
| bswartz | if any instance of a share is in an error state, then the share should be in an error state | 18:51 |
| *** xyang_ has quit IRC | 18:51 | |
| bswartz | the existing code is just dangerous because it allowed error instances to exist without preventing any further actions | 18:51 |
| bswartz | second, when migrations fail, we probably should not be setting the source instance back to "available" | 18:52 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support https://review.openstack.org/345526 | 18:52 |
| ganso | bswartz: why not? | 18:53 |
| bswartz | in many cases that might be the right thing to do, but as more vendors implement driver-assisted migraiton it might become a significant burden to always ensure that the source side is always something you can revert back to | 18:53 |
| ganso | bswartz: I thought that was the most important requirement | 18:53 |
| bswartz | a more generic approach is to allow everything to go into an error state when the migration fails, so the admin can clean it up | 18:53 |
| ganso | bswartz: We decided that in a meeting | 18:53 |
| bswartz | are you talking about austin? or earlier or later? | 18:54 |
| ganso | around newton midcycle probably | 18:54 |
| bswartz | before we talked about driver-assisted migration and we were focused on the fallback approach, that was true | 18:54 |
| bswartz | the fallback approach (which we write and own) should never corrupt your original copy | 18:55 |
| bswartz | but when drivers are doing this themselves, who knows how things might work internally? | 18:55 |
| bswartz | they could make state changes which are hard or impossible to revert | 18:55 |
| bswartz | do we want to say that such approaches are automatically invalid? | 18:55 |
| ganso | bswartz: if we follow this road, it may end up that the driver-assisted migration has risks of losing data | 18:56 |
| *** yumiriam has quit IRC | 18:56 | |
| bswartz | ganso: I don't think we'd go that far | 18:56 |
| ganso | bswartz: if we remove this requirement, I think we woudl | 18:56 |
| bswartz | you could end up with a driver-assisted migration that results in loss of access to your data until an admin is able to clean up a failure | 18:56 |
| *** lpetrut has quit IRC | 18:57 | |
| *** xyang_ has joined #openstack-manila | 18:57 | |
| bswartz | the theory is that admins are able to recover when migrations fail | 18:57 |
| bswartz | using some out-of-band operations and a reset-state on manila | 18:57 |
| ganso | bswartz: are you talking about failures in migration-complete or any point in migration? | 18:57 |
| bswartz | mostly the complete case | 18:57 |
| ganso | bswartz: ok | 18:58 |
| bswartz | because migrations are supposed to be cancel-able before then | 18:58 |
| ganso | bswartz: yes, exactly | 18:58 |
| bswartz | but I haven't thought through all the corner cases | 18:58 |
| bswartz | anyways I think it's worth considering the possibilty of both the migrating_from and the migrating_to instances going to a migration error state when something goes wrong | 18:59 |
| bswartz | the third issue was the most important | 18:59 |
| *** ianychoi has quit IRC | 19:00 | |
| ganso | bswartz: it will be a little bit harder to identify which is the source and which is the destination, only admin will be able to identify through the 'host' field... the code always checks for these statuses | 19:00 |
| ganso | bswartz: the code will not be able to distinguish between source and destination anymore | 19:00 |
| bswartz | gouthamr explained how the ShareSnapshot.instance() method works today and has a different idea of what we should do for migrating snapshots | 19:00 |
| bswartz | ganso: that's a good point | 19:01 |
| bswartz | ganso: but the admin will have the actual backends to look at, and the logs to | 19:01 |
| bswartz | he'll be able to figure it out | 19:01 |
| bswartz | besides it might be a moot point -- in the case of a migration, either side might be the "good" side after a failure | 19:02 |
| bswartz | it's equally likely that the admin will want to recover the source vs the destination | 19:02 |
| ganso | bswartz: during migration-complete, most likely yes | 19:02 |
| ganso | bswartz: what's the idea for ShareSnapshot.instance() ? | 19:03 |
| openstackgerrit | Vitaliy Levitski proposed openstack/manila-ui: Add MapRFS protocol https://review.openstack.org/421884 | 19:04 |
| bswartz | gouthamr pointed out that for replicated shares, the snapshot instance that should be exposed depends on the share instances states more than the snapshot instance states | 19:04 |
| bswartz | no matter what the status of the snapshots instances are, we want to expose the snapshot instance of the active share instance | 19:05 |
| bswartz | I think the same applies to migrating snapshots | 19:05 |
| ganso | bswartz: yes, but at this moment we do not support migration with replicas | 19:05 |
| bswartz | if the share is migrating, we should find the snapshot instance attached to the source instance always | 19:05 |
| gouthamr | in this case, the source of the migration | 19:05 |
| ganso | bswartz: yes I already do that | 19:05 |
| *** dustins has quit IRC | 19:06 | |
| gouthamr | ganso: can we do that without the sorting? | 19:06 |
| bswartz | but it should not be based on snapshot status | 19:06 |
| ganso | bswartz: within the code, it is no problem, the only concern was what the user sees when issues "manila list" or "manila snapshot-list" commands | 19:06 |
| bswartz | it should be based on share status | 19:06 |
| ganso | gouthamr: the sorting is for listing | 19:06 |
| gouthamr | ganso: we want to deliberately pick only the migrating instance's snapshot instance | 19:06 |
| gouthamr | ganso: listing at the API? | 19:06 |
| ganso | gouthamr: yes | 19:06 |
| openstackgerrit | Alyson proposed openstack/python-manilaclient: Add mountable snapshots support to manila client https://review.openstack.org/345625 | 19:06 |
| ganso | gouthamr: only that | 19:06 |
| gouthamr | ganso: hmmm, can we specifically pick the one we're interested in, by just looking at the share instance statuses? | 19:07 |
| bswartz | the point is that this model code gets used all over the place, not just for the list commands | 19:07 |
| ganso | gouthamr: we would need to change snapshot list view logic | 19:07 |
| ganso | bswartz: there are other places, not only listing or migration | 19:08 |
| ganso | bswartz: like, let's say we are migrating | 19:08 |
| ganso | bswartz: and user wants to create a share from the snapshot that is being migrated | 19:08 |
| ganso | bswartz: I believe that currently there is nothing preventing that, and why would we right? | 19:08 |
| ganso | bswartz: so, create share from snapshot API does snapshot.instance... totally unaware that there are more than one instance | 19:09 |
| bswartz | ganso: we probably shouldn't allow that | 19:09 |
| bswartz | there's a built in race condition there | 19:09 |
| bswartz | the snapshot creation will race against the migration competing | 19:09 |
| ganso | bswartz: well that was an example, but how about access rules | 19:09 |
| bswartz | err the share creation, not snapshot | 19:10 |
| bswartz | s/competing/completing/ | 19:10 |
| *** dustins has joined #openstack-manila | 19:10 | |
| gouthamr | ganso: for access rules, we will disallow adding/removing rules if ANY share instance is in an invalid state | 19:10 |
| ganso | gouthamr: MIGRATING will be an invalid state? | 19:11 |
| *** furlongm_ has quit IRC | 19:12 | |
| *** furlongm_ has joined #openstack-manila | 19:12 | |
| bswartz | ganso: I like the idea of blocking everything by default, until we know it's safe for 2 operations to proceed in parallel | 19:12 |
| bswartz | not being rigorous about this stuff is how the code got riddled with race conditions | 19:12 |
| ganso | bswartz: what is the advantage of the share being accessible in readonly or writable if the user cannot add rules? | 19:14 |
| bswartz | ganso: there's a difference between the share being accessible for I/O purposes and the share being accessible through manila for management purposes | 19:15 |
| gouthamr | ganso: there's a huge advantage in not disrupting existing clients | 19:15 |
| bswartz | all the manila services can be taken offline, and we still expect NFS operation to existing shares to continue without problems | 19:15 |
| *** xyang_ has quit IRC | 19:17 | |
| *** cdelatte has joined #openstack-manila | 19:18 | |
| *** xyang_ has joined #openstack-manila | 19:18 | |
| ganso | bswartz: ok so we are agreeing that we will change things again, MIGRATING status will block everything else, is that right? | 19:19 |
| ganso | gouthamr: If that's so, I gotta update the cast_to_readonly patch | 19:19 |
| *** ociuhandu has joined #openstack-manila | 19:19 | |
| *** xyang_ has quit IRC | 19:20 | |
| bswartz | ganso: that seems like a safer default to me | 19:20 |
| ganso | bswartz: ok | 19:20 |
| bswartz | unless we're going to spend the time to look at operations which should be safe to do at the same time and make sure we have all the races covered | 19:20 |
| bswartz | the interesting cases are parellel operations that occur right before or after the migration starts or completes | 19:21 |
| ganso | bswartz: yes | 19:21 |
| ganso | bswartz: some operations are ok for the most part | 19:21 |
| ganso | bswartz: unless we increase the granularity of the statuses | 19:21 |
| bswartz | I agree a create-share-from-snapshot in the middle of a long migration should be safe | 19:21 |
| ganso | with several -ING statuses | 19:21 |
| bswartz | similarly changing access | 19:22 |
| bswartz | but they're not safe right around the transitions and unless we have good locks we simply shouldn't allow them | 19:22 |
| ganso | bswartz: but we don't have time to do that right now :\ | 19:22 |
| bswartz | right | 19:22 |
| ganso | bswartz: also, our locks are worthless in a HA environment without Tooz | 19:23 |
| *** timcl has quit IRC | 19:24 | |
| *** mtanino has joined #openstack-manila | 19:25 | |
| bswartz | merge the tooz patch! :-D | 19:25 |
| ganso | bswartz: so the less possibility for race conditions, the better | 19:25 |
| bswartz | agreed | 19:25 |
| *** absubram has joined #openstack-manila | 19:25 | |
| ganso | bswartz: I haven't tested it myself, I trust the judgement of the other cores working on it :) | 19:25 |
| *** furlongm_ has quit IRC | 19:26 | |
| ganso | bswartz: about the ordering change | 19:27 |
| ganso | bswartz: are we going to fix that in a separate patch? | 19:27 |
| *** timcl has joined #openstack-manila | 19:27 | |
| bswartz | for share instances? | 19:27 |
| ganso | bswartz: ya | 19:27 |
| bswartz | yes don't worry about fixing that in your patch | 19:27 |
| bswartz | we should bug it and deal with it after FF | 19:27 |
| ganso | bswartz: I think it is better, it has been broken for a while, it will be confusing for a migration patch to change this behavior, this is a line the migration patch is not currently changing | 19:27 |
| ganso | bswartz: ok | 19:28 |
| bswartz | for your patch I'd like to see the snapshot instance returned always be the migration source when a migration is in progress | 19:28 |
| bswartz | if that's not too much effort to implement | 19:28 |
| *** furlongm_ has joined #openstack-manila | 19:29 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Tooz integration https://review.openstack.org/318336 | 19:29 |
| ganso | bswartz: so that's what we were discussing before you head out for lunch | 19:30 |
| ganso | bswartz: gouthamr advised to put the migrating status before available, but still after deleting and error | 19:30 |
| bswartz | I was saying that method should be blind to the snapshot status and look only at the share status | 19:32 |
| *** openstackgerrit has quit IRC | 19:33 | |
| ganso | bswartz: that's a change to the view, not migration at all | 19:33 |
| bswartz | it's a change to the model object, which is used all over the place unless I'm mistaken | 19:34 |
| ganso | bswartz: I believe it does not look at the share at all currently | 19:34 |
| bswartz | ganso: https://github.com/openstack/manila/blob/master/manila/db/sqlalchemy/models.py#L626 | 19:35 |
| bswartz | the existing code uses that filter function on the share instances | 19:36 |
| ganso | bswartz: only for replicas | 19:36 |
| bswartz | I'm imagining a similar approach for migration instances | 19:37 |
| ganso | bswartz: status == constants.MIGRATING | 19:37 |
| *** david-lyle has joined #openstack-manila | 19:48 | |
| ganso | bswartz: the migration code guarantees that the share instance with migrating status has its respective snapshot instance set to migrating as well | 19:54 |
| ganso | bswartz: I am coding the logic to check for the share instance status but it seems unnecessary | 19:54 |
| *** xyang_ has joined #openstack-manila | 19:57 | |
| *** xyang_ has quit IRC | 19:58 | |
| *** absubram has quit IRC | 20:06 | |
| *** jprovazn has quit IRC | 20:09 | |
| bswartz | ganso: since this model sits underneath everything I think it's important to establish good patterns | 20:10 |
| bswartz | I'm not sure what the next usage of multiple instances will be | 20:10 |
| *** openstackgerrit has joined #openstack-manila | 20:11 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Refactor Access Rules APIs https://review.openstack.org/369668 | 20:11 |
| *** darrenc_ is now known as darrenc | 20:15 | |
| *** xyang_ has joined #openstack-manila | 20:16 | |
| *** zengyingzhe__ has joined #openstack-manila | 20:25 | |
| *** zengyingzhe_ has quit IRC | 20:25 | |
| openstackgerrit | Yogesh proposed openstack/manila: Improve test coverage for share migration https://review.openstack.org/418559 | 20:28 |
| *** lpetrut has joined #openstack-manila | 20:29 | |
| ganso | bswartz: ok I added the code, will submit in a few minutes | 20:30 |
| *** xyang_ has quit IRC | 20:33 | |
| *** xyang_ has joined #openstack-manila | 20:46 | |
| *** pablo|500| has quit IRC | 20:57 | |
| *** kaisers has joined #openstack-manila | 20:58 | |
| *** timcl has quit IRC | 20:59 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Share Migration Ocata Improvements https://review.openstack.org/406305 | 21:07 |
| *** kaisers has quit IRC | 21:07 | |
| *** nkrinner_afk has quit IRC | 21:15 | |
| *** absubram has joined #openstack-manila | 21:15 | |
| *** nkrinner_afk has joined #openstack-manila | 21:16 | |
| *** zhonghua has joined #openstack-manila | 21:23 | |
| gouthamr | xyang_ bswartz: The tooz patch is ready for your re-review https://review.openstack.org/#/c/318336/ | 21:25 |
| *** zhonghua2 has quit IRC | 21:26 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support https://review.openstack.org/345526 | 21:26 |
| *** eharney has quit IRC | 21:28 | |
| openstackgerrit | Yogesh proposed openstack/manila: Improve test coverage for share migration https://review.openstack.org/418559 | 21:32 |
| *** david-lyle has quit IRC | 21:35 | |
| xyang_ | gouthamr: ok | 21:36 |
| * gouthamr xyang_: thank you. | 21:37 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add cast_rules_to_readonly to share instances https://review.openstack.org/419163 | 21:37 |
| *** david-lyle has joined #openstack-manila | 21:39 | |
| *** david-lyle has quit IRC | 21:39 | |
| *** david-lyle has joined #openstack-manila | 21:39 | |
| *** Yogi1 has quit IRC | 21:40 | |
| *** xyang_ has quit IRC | 21:47 | |
| *** xyang_ has joined #openstack-manila | 21:48 | |
| *** xyang_ has quit IRC | 21:49 | |
| *** xyang_ has joined #openstack-manila | 21:49 | |
| *** cknight has quit IRC | 21:51 | |
| *** cknight has joined #openstack-manila | 21:51 | |
| *** dustins has quit IRC | 21:52 | |
| *** cdelatte has quit IRC | 21:58 | |
| *** lpetrut has quit IRC | 22:03 | |
| *** JoseMello has quit IRC | 22:05 | |
| *** alyson_ has quit IRC | 22:07 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Share Migration Ocata Improvements https://review.openstack.org/406305 | 22:09 |
| *** adrianofr_ has quit IRC | 22:29 | |
| openstackgerrit | Helen Walsh proposed openstack/manila: VMAX manila plugin - Support for VMAX in Manila https://review.openstack.org/404859 | 22:30 |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Improve test coverage for share migration https://review.openstack.org/418559 | 22:45 |
| *** gouthamr has quit IRC | 22:45 | |
| *** ianychoi has joined #openstack-manila | 22:49 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add mountable snapshots support https://review.openstack.org/345526 | 22:50 |
| *** gouthamr has joined #openstack-manila | 23:02 | |
| *** david-lyle has quit IRC | 23:03 | |
| *** tommylikehu_ has joined #openstack-manila | 23:03 | |
| *** xyang_ has quit IRC | 23:04 | |
| *** tommylikehu_ has quit IRC | 23:05 | |
| *** absubram has quit IRC | 23:05 | |
| *** absubram has joined #openstack-manila | 23:09 | |
| *** makowals has quit IRC | 23:17 | |
| *** absubram has quit IRC | 23:23 | |
| *** cknight has quit IRC | 23:24 | |
| *** makowals has joined #openstack-manila | 23:55 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!