15:01:34 #startmeeting manila 15:01:34 Meeting started Thu Nov 12 15:01:34 2020 UTC and is due to finish in 60 minutes. The chair is gouthamr. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:36 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:01:38 The meeting name has been set to 'manila' 15:01:50 o/ 15:01:57 courtesy ping: ganso vkmc dviroel lseki tbarron andrebeltrami felipe_rodrigues 15:01:59 hi 15:02:00 hi 15:02:03 hi 15:02:04 o/ 15:02:04 o/ 15:02:23 hello o/ 15:02:58 thanks for joining, lets get going 15:03:03 #topic Announcements 15:03:38 first off, vhari has posted a notification to openstack-discuss regarding next week's bug squash event 15:03:52 #link http://lists.openstack.org/pipermail/openstack-discuss/2020-November/018629.html (bug squash on 19th Nov 2020, no IRC meeting) 15:04:43 we will be hosting a sync-up call in the place of this irc meeting on that day 15:05:03 so please bookmark the call URL: 15:05:14 #link https://meetpad.opendev.org/ManilaW-ReleaseBugSquash (Jitsi room for the bugsquash) 15:05:50 and do take a look at the list of bugs that have been put together - many of them are assigned, and have been looked at at some point in the past 15:06:03 #link https://ethercalc.openstack.org/birpr9a6bd0b (bug list for the bug squash) 15:07:10 vhari's added bugs here that have been ascertained to need api/db changes; something we were uncomfortable doing during the last bug squash event 15:07:39 so we can use the sync call to discuss any of these bugs, and brainstorm 15:08:14 i'm hoping you can all make it and we have a successful bug squash! :) 15:08:43 ty gouthamr . pls feel free to flag any other candidate bugs for the event 15:09:18 ^ +1, if there's some bug you're looking to tackle, please bring it (or add it to the ethecalc above) 15:09:29 +1 15:09:47 ty for planning this vhari :) 15:10:27 alright, any questions/concerns? 15:10:32 gouthamr, yw .. excited to squash some bugs together :) 15:11:06 great! 15:11:29 next up, we have an OpenStack community meeting today, at 1600 UTC 15:12:01 #link http://lists.openstack.org/pipermail/openstack-discuss/2020-November/018646.html (OpenStack Community Meeting on Nov 12, 1600 UTC) 15:12:34 we should expect to see some Victoria project updates, and have a general Q&A session 15:12:52 so please join in if you can! 15:13:48 and finally, maaritamm has posted a prioritized list of OSC commands that are left 15:13:59 #link http://lists.openstack.org/pipermail/openstack-discuss/2020-November/018633.html (Priorities for implementing the OSC share commands) 15:14:30 maaritamm++ 15:15:29 i'd like for us to take a look at this list, and do share your views on the priority of commands for implementation 15:16:07 as we know this has been a multi-cycle effort for us, and the committer/reviewer bandwidth is limited - so having a priority hugely helps! 15:16:55 there are a few patches up for review already 15:17:08 :) 15:17:46 she's super fast! maaritamm++ 15:18:24 great, that's a wrap on announcements for today, anyone else got any 15:18:26 ? 15:19:04 alright, excuse me while i make an ad hoc topic 15:19:35 #topic No remaining IRC meetings in November 15:20:33 So, as mentioned prior, the next IRC meeting on 19th Nov has been canceled and we'll be meeting on a Jitsi call instead 15:21:33 the week after that is a holiday week, at least here in the US - and i wanted to ask if we should go ahead and take a break 15:22:19 (and possibly other countries) 15:22:55 are there any concerns with taking this break? 15:23:35 ++ for the break :) 15:24:24 ty carloss, the only thing i can think of is that milestone-1 is the week after 15:25:10 we have no release deliverables, but, it's a bug fix milestone 15:26:27 so if folks want to talk about bugs, we'd have to do that on #openstack-manila or the openstack-discuss mailing list 15:27:33 alright, lets take an AI 15:27:58 #action gouthamr will send an email to openstack-discuss canceling the two IRC meetings remaining in Nov 15:29:19 we've run through the agenda pretty much, lets move down our regular schedule 15:29:23 #topic Bugs (vhari) 15:29:31 back to you, vhari :) 15:29:55 gouthamr, we have a number of new bugs on our radar to triage today! 15:30:01 first off .. 15:30:06 #link https://bugs.launchpad.net/manila/+bug/1904015 15:30:07 Launchpad bug 1904015 in OpenStack Shared File Systems Service (Manila) "Manila overwrite existing Ceph users" [Undecided,New] 15:30:30 oh this one comes from the openstack-discuss list 15:30:38 #link http://lists.openstack.org/pipermail/openstack-discuss/2020-November/018684.html (Manila user overwriting existing Ceph users) 15:30:40 ack .. 15:31:33 interesting 15:33:13 and a pretty ugly bug, since a user can break other services 15:33:36 if they can discover the cephx user used to configure these services 15:34:20 vhari: i can investigate this bug and respond 15:34:51 gouthamr, sounds good .. 15:35:07 i think we can set this to medium, and bump up the severity if we can confirm it does indeed mess with the infrastructure 15:35:44 sounds good 15:35:59 not sure if this should be handled in ceph side or in manila side 15:36:56 good point, i think we do some evictions and denials in the driver today based on things we know shouldn't work 15:37:14 i'll do some more digging and we can talk, vkmc 15:37:32 gouthamr, cool 15:37:43 cool, moving on 15:37:55 #link https://bugs.launchpad.net/manila/+bug/1903922 15:37:56 Launchpad bug 1903922 in OpenStack Shared File Systems Service (Manila) "Generic driver fails CIFS online extension test " [Undecided,New] 15:38:25 ah i reported this one, and the rca isn't complete 15:38:40 gouthamr, ack and you posted repro config 15:38:57 my bad on the timing here - i enabled cifs testing in the generic driver a while ago, and we merged the patch with this failure 15:39:35 right now, it also blocks upgrading the CI job to focal fossa 15:40:09 #link https://review.opendev.org/762218/ ([ci] Switch generic driver job to focal fossa) 15:42:19 i can keep this in my backlog and work on it opportunistically since we don't have any dedicated generic driver maintainers 15:42:39 but, if anyone's willing to help - i'd more than appreciate it! 15:44:28 share extensions in the generic driver are disruptive and clients may need to remount depending on how long cinder takes to perform the extension and if the client has a hard mount, and we're able to complete the whole transaction within their stale mount timeout 15:45:11 s/are disruptive/can be disruptive 15:45:35 so we might have a workaround - i'll test it and post it on the bug 15:46:35 for triage though, i think this is a medium and it's not user reported - likely this is something that breaks only on our CI 15:47:55 gouthamr, done .. 15:48:00 ty vhari 15:48:09 we have time to look at one more new bug .. 15:48:13 #link https://bugs.launchpad.net/manila/+bug/1900752 15:48:14 Launchpad bug 1900752 in OpenStack Shared File Systems Service (Manila) "cifs without security service error not helpful" [Undecided,New] 15:48:58 hmmm, this is a tricky one 15:49:20 it's really up to the drivers if CIFS shares support security services 15:49:45 Generic, Windows and HPE for instance can work without it (Generic doesn't support security services yet) 15:50:16 ONTAP and the Dell/EMC drivers *need* Active Directory 15:50:21 agree though that the error message should be more helpful 15:50:43 vkmc ++ 15:50:53 "create: An unknown error occurred." is very... shady 15:50:55 def agree that this can be signalled with a user message, maybe we can make it generic and have the drivers raise a particular exception 15:51:06 not sure where it is raised 15:51:09 * gouthamr overuses the word generic 15:51:20 and if we have enough info to craft a better user message for this error 15:51:56 i don't think we currently do - backends seem to be raising a ShareBackendException of sorts and the share manager is logging the information 15:54:01 we could prevent share creation if we had this info prior - but, the api service doesn't know which specific backend's going to get the request 15:57:01 anyone willing to take a look at this and see what can be done? since it affects netapp, dell/emc and possibly other vendor backends afaict, i think this should be a medium priority bug at least 15:57:42 I can take a look and find some options 15:57:55 ty carloss 15:58:03 then I can bring the options to the table and we can discuss them 15:58:12 yw :) 15:58:16 ++ 15:58:17 carloss++ 15:58:20 gouthamr, all caught up triaging new bugs ... thx all for participating :) 15:58:29 thank you vhari! 15:58:35 #topic Open Discussion 15:58:49 we have all of two minutes :) 15:59:01 any final thoughts? 15:59:51 I'm thought-less 15:59:58 :D alright, thank you all for attending - hope you can make it to the community meeting that's happening next, and the manila bugsquash event next week 16:00:11 lets take further thoughts to #openstack-manila 16:00:14 #endmeeting