| gouthamr | sticker: if you're still around, no reason that seems obvious | 00:01 |
|---|---|---|
| sticker | cool, have submitted a bug report and will submit a fix | 00:02 |
| sticker | https://bugs.launchpad.net/manila/+bug/1715768 | 00:02 |
| openstack | Launchpad bug 1715768 in Manila "Unable to use dollar sign in usernames for access" [Undecided,New] - Assigned to Daniel Russell (danielr-2) | 00:02 |
| gouthamr | sticker: nice, thanks! | 00:02 |
| openstackgerrit | Daniel Russell proposed openstack/manila master: Allows the use of dollar sign in usernames https://review.openstack.org/501903 | 00:09 |
| openstackgerrit | Daniel Russell proposed openstack/python-manilaclient master: Allows the use of dollar sign in usernames https://review.openstack.org/501905 | 00:17 |
| openstackgerrit | Daniel Russell proposed openstack/manila master: Allows the use of dollar sign in usernames https://review.openstack.org/501906 | 00:21 |
| *** baojg has quit IRC | 00:26 | |
| *** harlowja has quit IRC | 01:11 | |
| zhongjun | gouthamr:ping | 01:12 |
| gouthamr | zhongjun: pong | 01:13 |
| bswartz | zhongjun: get your auth fixed yet? I still haven't seen a PM | 01:13 |
| zhongjun | gouthamr: We can only get metadata info by API() method, and get metadata from db https://github.com/openstack/manila/blob/master/manila/share/drivers/maprfs/maprfs_native.py#L117 | 01:13 |
| zhongjun | bswartz: I seems not work | 01:14 |
| bswartz | :-( | 01:14 |
| gouthamr | zhongjun: yes, like share type extra specs | 01:15 |
| zhongjun | gouthamr: Do you know why we doesn't add metadata info in driver apis? | 01:15 |
| gouthamr | zhongjun: you mean as part of the call from the manager to the share driver? | 01:16 |
| gouthamr | zhongjun: probably because you don't know when you need it and when you don't.. i recall this discussion regarding share type extra specs.. don't see a problem with this pattern though. | 01:17 |
| zhongjun | gouthamr: yes | 01:17 |
| gouthamr | zhongjun: not a lot of drivers acknowledge share metadata.. also, users can change metadata at any time and there's no "update" call to the share drivers | 01:19 |
| gouthamr | called it out to the maprfs folks when they wrote their driver, but they said they'll deal with documenting when/why "_name" would make sense when using MAPRFS shares | 01:20 |
| zhongjun | gouthamr: It just look a little werid. The driver could call common method, but in this case the driver call api() method | 01:22 |
| zhongjun | s/werid/weird | 01:23 |
| gouthamr | zhongjun: it's a wrapper around a DB call.. like "get_share_type_extra_specs" being used in many drivers | 01:25 |
| zhongjun | gouthamr: yeah, it's a wrapper around a DB call, we could just put it in a common place if most of driver need to call it, like "extract_host" being used in many drivers | 01:31 |
| zhongjun | gouthamr: but it is not a big deal | 01:34 |
| gouthamr | zhongjun: +1 | 01:36 |
| zhongjun | gouthamr: Could we quickly check another confuse for me | 01:37 |
| gouthamr | zhongjun: sure.. | 01:37 |
| zhongjun | gouthamr: After I create a replica, the replication id is equal to share instance id. If we delete all the share replica, it means we will delete all share instances | 01:38 |
| zhongjun | gouthamr: then the share doesn't work? | 01:38 |
| gouthamr | zhongjun: you can't delete the last share instance without deleting the share as well... | 01:39 |
| zhongjun | gouthamr: If we create a share replica, then we can not allowed to delete all share replications? | 01:39 |
| gouthamr | zhongjun: yes, DELETE /share-replicas will only work as long as there's atleast one active replica | 01:40 |
| zhongjun | gouthamr: I used 'share-replica-delete --force' , then I deleted the share instance | 01:40 |
| gouthamr | zhongjun: same with POST /share-replicas/{id}/action {'force-delete': null} <--- you can't delete the last active replica with this API | 01:41 |
| zhongjun | gouthamr: so, if we create a share replica for a share, it means we can never delete it | 01:42 |
| gouthamr | however, you can mess things up if you went after your replicas with the /share_instances API | 01:42 |
| gouthamr | zhongjun: you can always DELETE /shares/{id} | 01:42 |
| zhongjun | gouthamr: It could be mess things, but I just think whether it could be weird for the user that he can not delete all share replications that he created and always keep a share replication? | 01:46 |
| gouthamr | zhongjun: when you create a share with a share type that supports replication, you always have a "replica... but that doesn't mean | 01:47 |
| gouthamr | your share is being replicated yet | 01:47 |
| gouthamr | zhongjun: look at the share's "has_replicas" field after your share has been created | 01:49 |
| gouthamr | zhongjun: you can then create 1 or more secondary replicas to your share and delete them at will, you can't delete the share itself with "manila share-replica-delete" | 01:49 |
| zhongjun | gouthamr: we create a share with a share type that supports replication, it means we have abilities to create a replica, it seems doesn't mean we always have a "replica | 01:49 |
| gouthamr | zhongjun: true.. you've only created a share that can be replicated.. but it isn't unless you create share replicas | 01:50 |
| zhongjun | gouthamr: After I tested " POST /share-replicas/{id}/action {'force-delete': null} <--- you can't delete the last active replica with this API" I can delete the last non-sync replica | 01:52 |
| gouthamr | zhongjun: what do you mean by non-sync? | 01:53 |
| zhongjun | gouthamr: I mean it can be deleted if the status is not active | 01:54 |
| gouthamr | zhongjun: mostly correct.. you can delete any non-active replica with "manila share-replica-delete" | 01:55 |
| gouthamr | zhongjun: slightly different only in case of "writable" replication, where all replicas are "active" and "manila share-replica-delete" will not work for the last one | 01:56 |
| zhongjun | gouthamr: so you means normal at least the one replica status is active, then we shouldn't delete it. If the share replica status is not active, and we delete this share replica, then the relative share instance also be deleted,In this case, it is correct. | 02:02 |
| gouthamr | correct | 02:02 |
| gouthamr | if you have no replicas, you can delete your share. if you have replicas, you cannot delete your share until you delete all your replicas. Sort of like Share and Snapshots - you can't delete your share until you delete all the snapshots. | 02:04 |
| *** markstur has quit IRC | 02:05 | |
| zhongjun | gouthamr: but we have two status in share replica, my case is: 1. create a share 2 create a share replica (status: available, Replica State: active) 3 change the share replica status (status: available, Replica State: out_of_sync) 4: delete all the share replica The results is we delete the share replica and delete the share instance, and also the share has been deleted. | 02:16 |
| gouthamr | 1) create a share (status: available, replica_state: active, Share has_replicas = False) | 02:17 |
| gouthamr | 2) create a replica (status: available, replica_state: out_of_sync/in_sync, Share has_replicas = True) | 02:18 |
| zhongjun | gouthamr: like share and snapshot, we delete the snapshot, but we don't delete the share at the same time | 02:18 |
| gouthamr | what do you mean by (3)? | 02:19 |
| zhongjun | gouthamr: the share replica status has been changed to out_of_sync | 02:20 |
| gouthamr | zhongjun: with share-replica-reset-state? | 02:20 |
| zhongjun | gouthamr: yeah | 02:20 |
| gouthamr | share-replica-reset-replica-state* | 02:20 |
| zhongjun | gouthamr: or it could be updated by driver | 02:21 |
| zhongjun | gouthamr: or it could be updated by driver when we run promote or something else | 02:21 |
| gouthamr | zhongjun: don't see how drivers can change from "active" to "out_of_sync" unless it is during a promotion | 02:22 |
| gouthamr | zhongjun: yes, the secondary becomes "active" after a promotion, and cannot be deleted | 02:23 |
| *** markstur has joined #openstack-manila | 02:25 | |
| *** markstur has quit IRC | 02:25 | |
| zhongjun | gouthamr: https://bugs.launchpad.net/manila/+bug/1715783 | 02:27 |
| openstack | Launchpad bug 1715783 in Manila "Share replica problem" [Undecided,New] | 02:27 |
| zhongjun | gouthamr: steps | 02:27 |
| gouthamr | zhongjun: thanks, i understand what you mean by (3) | 02:29 |
| gouthamr | zhongjun: what you're doing what shouldn't be done.. reset-replica-state is an admin API only to assist in resetting the state when administrators have corrected issues on the backend.. it's helpful in case of a failed promotion | 02:32 |
| gouthamr | zhongjun: unsure how we can prevent misuse | 02:32 |
| gouthamr | :) | 02:32 |
| gouthamr | perhaps fail if there's only one replica | 02:33 |
| gouthamr | i.e, only allow setting it to "active" | 02:33 |
| *** scottda_ has quit IRC | 02:37 | |
| zhongjun | gouthamr : There could build a way to warn the end user: "notice ~~ your share will be deleted " | 02:41 |
| *** catintheroof has joined #openstack-manila | 02:48 | |
| *** catinthe_ has joined #openstack-manila | 02:50 | |
| *** catintheroof has quit IRC | 02:51 | |
| openstackgerrit | zhongjun proposed openstack/manila-specs master: Add count info in list response https://review.openstack.org/501934 | 03:21 |
| *** catinthe_ has quit IRC | 03:56 | |
| *** gouthamr has quit IRC | 04:19 | |
| *** rejy has joined #openstack-manila | 05:09 | |
| *** gaurangt has quit IRC | 05:17 | |
| *** gaurangt has joined #openstack-manila | 05:19 | |
| *** dims has quit IRC | 05:54 | |
| *** harlowja has joined #openstack-manila | 06:19 | |
| *** jprovazn has joined #openstack-manila | 06:21 | |
| *** harlowja has quit IRC | 06:40 | |
| *** sticker has quit IRC | 06:43 | |
| *** dims has joined #openstack-manila | 06:51 | |
| *** gcb has quit IRC | 07:08 | |
| openstackgerrit | zhongjun proposed openstack/manila master: [Doc] Fix doc description for consistency group https://review.openstack.org/501198 | 07:12 |
| openstackgerrit | zhongjun proposed openstack/manila master: [Doc] Delete doc description for consistency group https://review.openstack.org/501198 | 07:16 |
| openstackgerrit | zhongjun proposed openstack/manila master: [Doc] Delete consistency group in doc https://review.openstack.org/501198 | 07:21 |
| *** chlong has joined #openstack-manila | 07:36 | |
| *** chlong has quit IRC | 07:45 | |
| *** chlong has joined #openstack-manila | 08:25 | |
| *** chlong has quit IRC | 08:35 | |
| *** chlong has joined #openstack-manila | 08:51 | |
| *** chlong_ has joined #openstack-manila | 08:54 | |
| *** chlong has quit IRC | 08:58 | |
| *** chlong_ has quit IRC | 09:04 | |
| *** rraja has joined #openstack-manila | 09:17 | |
| *** chlong has joined #openstack-manila | 09:21 | |
| *** chlong has quit IRC | 09:24 | |
| *** chlong has joined #openstack-manila | 09:25 | |
| *** baojg has joined #openstack-manila | 09:41 | |
| *** jprovazn has quit IRC | 09:44 | |
| *** chlong has quit IRC | 09:55 | |
| -openstackstatus- NOTICE: Our CI systems experience a hickup, no new jobs are started. Please stay tuned and wait untils this resolved. | 10:47 | |
| *** baojg has quit IRC | 11:14 | |
| *** jprovazn has joined #openstack-manila | 11:34 | |
| *** ganso has joined #openstack-manila | 11:35 | |
| *** baojg has joined #openstack-manila | 11:49 | |
| *** jprovazn has quit IRC | 11:55 | |
| *** jprovazn has joined #openstack-manila | 11:58 | |
| *** jungleboyj has quit IRC | 12:06 | |
| *** eharney has quit IRC | 12:46 | |
| *** catintheroof has joined #openstack-manila | 13:26 | |
| *** gouthamr has joined #openstack-manila | 13:27 | |
| *** eharney has joined #openstack-manila | 13:34 | |
| *** jprovazn has quit IRC | 13:36 | |
| *** cknight has joined #openstack-manila | 13:37 | |
| *** xyang1 has joined #openstack-manila | 13:40 | |
| *** catinthe_ has joined #openstack-manila | 13:41 | |
| *** catintheroof has quit IRC | 13:44 | |
| *** jprovazn has joined #openstack-manila | 13:45 | |
| -openstackstatus- NOTICE: nodepool issue related to bad images has been resolved, builds should be coming back online soon. Restarted gerrit due to reasons. Happy Friday. | 13:46 | |
| *** scottda_ has joined #openstack-manila | 13:48 | |
| *** rraja has quit IRC | 13:57 | |
| *** zengyingzhe_ has quit IRC | 14:07 | |
| *** zengyingzhe_ has joined #openstack-manila | 14:07 | |
| *** rraja has joined #openstack-manila | 14:13 | |
| *** catintheroof has joined #openstack-manila | 14:22 | |
| *** catinthe_ has quit IRC | 14:25 | |
| *** jprovazn has quit IRC | 14:35 | |
| *** rraja has quit IRC | 14:40 | |
| *** rraja has joined #openstack-manila | 14:52 | |
| *** markstur has joined #openstack-manila | 14:56 | |
| *** cknight has quit IRC | 14:57 | |
| *** td has quit IRC | 15:07 | |
| *** cknight has joined #openstack-manila | 15:10 | |
| *** jungleboyj has joined #openstack-manila | 15:15 | |
| *** rraja has quit IRC | 15:26 | |
| *** markstur has quit IRC | 16:03 | |
| *** rejy has quit IRC | 16:25 | |
| *** dustins has joined #openstack-manila | 16:35 | |
| *** harlowja has joined #openstack-manila | 17:00 | |
| *** jmlowe has quit IRC | 17:09 | |
| *** jmlowe has joined #openstack-manila | 17:10 | |
| *** eharney has quit IRC | 18:31 | |
| *** eharney has joined #openstack-manila | 18:44 | |
| *** tinwood has quit IRC | 19:05 | |
| *** tinwood has joined #openstack-manila | 19:07 | |
| openstackgerrit | Tom Barron proposed openstack/manila master: Fix test_rpc_consumer_isolation for oslo.messaging 5.31.0 https://review.openstack.org/502173 | 20:24 |
| tbarron | https://bugs.launchpad.net/cinder/+bug/1715470 | 20:25 |
| openstack | Launchpad bug 1715470 in OpenStack Compute (nova) "test_rpc_consumer_isolation fails with oslo.messaging 5.31.0" [Medium,Fix released] - Assigned to Matt Riedemann (mriedem) | 20:25 |
| tbarron | bswartz: xyang1 toabctl gouthamr ^^^ pulled in that fix as our unit tests also fail w/o it | 20:25 |
| tbarron | ganso: vponomaryov cknight ^^ | 20:26 |
| bswartz | tbarron: that looks nice and sketchy -- I like it | 20:27 |
| tbarron | bswartz: I'm just copying the black magic | 20:27 |
| tbarron | can't say I like it. | 20:28 |
| tbarron | nova heritage | 20:29 |
| bswartz | I just mean it seems like a very open-stacky way to do things | 20:29 |
| tbarron | yeah | 20:30 |
| bswartz | push a change, break everyone, somebody fixes it with a hack, and everyone cargo cults the hack | 20:30 |
| tbarron | rofl | 20:32 |
| *** eharney has quit IRC | 20:49 | |
| *** gouthamr has quit IRC | 20:56 | |
| *** dustins has quit IRC | 21:22 | |
| *** gouthamr has joined #openstack-manila | 21:23 | |
| gouthamr | never call me ever. but it is okay to call me, sometimes, maybe. | 21:29 |
| gouthamr | lol https://review.openstack.org/#/c/502173/1/manila/tests/test_test.py | 21:29 |
| *** ying_zuo has left #openstack-manila | 21:46 | |
| *** xyang1 has quit IRC | 21:53 | |
| tbarron | but I'll ignore you if you do | 22:05 |
| *** catintheroof has quit IRC | 22:08 | |
| *** cknight has quit IRC | 22:24 | |
| *** baojg has quit IRC | 22:29 | |
| *** baojg has joined #openstack-manila | 22:30 | |
| *** ganso has quit IRC | 22:44 | |
| *** baojg has quit IRC | 22:53 | |
| *** baojg has joined #openstack-manila | 22:57 | |
| *** baojg has quit IRC | 23:05 | |
| *** baojg has joined #openstack-manila | 23:08 | |
| *** baojg has quit IRC | 23:16 | |
| *** baojg has joined #openstack-manila | 23:24 | |
| *** baojg has quit IRC | 23:40 | |
| *** baojg has joined #openstack-manila | 23:45 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!