04:00:50 #startmeeting masakari 04:00:51 Meeting started Tue Feb 28 04:00:50 2017 UTC and is due to finish in 60 minutes. The chair is samP. Information about MeetBot at http://wiki.debian.org/MeetBot. 04:00:52 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 04:00:54 The meeting name has been set to 'masakari' 04:01:13 Hi all, right after PTG.. 04:01:25 lets start with bugs... 04:01:39 #topic open bugs 04:01:48 # link: https://bugs.launchpad.net/masakari/+bug/1663513 04:01:48 Launchpad bug 1663513 in masakari "Masakari failed to rescue PAUSED instances" [Undecided,Confirmed] - Assigned to Dinesh Bhor (dinesh-bhor) 04:01:48 Hi all 04:02:28 Need feedback on Dinesh's comment on LP bug 04:03:24 # link: https://bugs.launchpad.net/masakari/+bug/1663513/comments/5 04:03:24 Launchpad bug 1663513 in masakari "Masakari failed to rescue PAUSED instances" [Undecided,Confirmed] - Assigned to Dinesh Bhor (dinesh-bhor) 04:04:31 I agree Dinesh opinion. I also Masakari should maintain the vm_state before recovery. 04:04:56 I agree that we need to maintain the consistency for VM state. Hoever rescue VM has same issue, right? 04:05:27 samP: yes 04:06:01 OK, then solution is, store the previous state of the VM and, return it to previous state after doing recovery 04:06:49 I mean, the solution should be a general one for all cases. 04:07:38 samP: At the time of instance recovery action, if the vm is in paused vm state, then what should be the final vm state after executing recovery action? 04:08:26 samP: this question is in context of the scenario highlighted in above LP bug 04:09:25 tpatil: I think we can not use reset-state api to set the state to PAUSED, right? 04:09:57 samP: yes, we can only set to error or active 04:11:31 for instance recovery action it should be, paused-> active -> paused 04:12:46 tpatil: IMO, final state should be paused 04:12:54 samP: OK, for all other cases, vm state will go into ACTIVE status 04:13:23 samP: for above LP bug, the final vm_state will be PAUSED 04:13:38 tpatil: yes. 04:13:51 samP: Thank you 04:13:52 I think PAUSED VM are going to lost their memory, and it not working before recovery, so is it good to remain it SHUTDOWN? 04:14:43 sagara: You have a good point there 04:15:39 sagara: correct. once we start the VM, we will lose the internal state of the VM 04:17:01 on the other hand, if shutdown it, it will not be the expected recovery 04:17:44 There should be some indication to the user that the instance has been recovered due to qemu process termination 04:19:09 if we keep the vm_state into PAUSED state 04:21:26 What would operators do if such problem occurs at present? 04:21:51 if the VM in PAUSED state, doing nothing would be a solution? 04:22:25 tpatil: I think currently we are not using that API, but let me confirm it 04:22:56 samP: please confirm it. 04:23:08 It might be nice to have some configuration. operator(admin) or user can specify VM recovery state, system default and per VM or per segment. 04:23:10 sagara: sure, I will do. 04:25:32 I will ask around for how ops handaling this issue, including sagara's point. 04:25:59 And, I will update the LP for this. 04:26:07 samP: Ok, please add your feedback on the LP bug 04:26:21 samP: thank you 04:26:49 #action samP Add FP to https://bugs.launchpad.net/masakari/+bug/1663513 04:26:49 Launchpad bug 1663513 in masakari "Masakari failed to rescue PAUSED instances" [High,Confirmed] - Assigned to Dinesh Bhor (dinesh-bhor) 04:27:18 Other bugs? 04:28:08 Takahara reported a bug 04:28:26 # link: https://bugs.launchpad.net/masakari/+bug/1667246 04:28:26 Launchpad bug 1667246 in masakari "When masakari-engine adds reserved_host to aggregate, 404 error occurs" [Undecided,Fix released] - Assigned to takahara.kengo (takahara.kengo) 04:29:14 Honjo and me, already reviewed this patch, and yesterday it merged. 04:30:03 sagara: thanks 04:30:50 samP: Do we need to back port this patch to stable/ocata branch? 04:31:16 tpatil: thatz what im thinking now 04:32:58 I think this is critical and better to back port to oacata. any objections? 04:33:05 samP, tpatil: I think it is high or critical bug. reserved host feature could not work without this patch, so we must do backport that stable/ocata. 04:34:14 if no objections, lets back port this to stable/ocata 04:34:44 samP: I agree to backport. 04:34:55 samP: sure, we will upload patch today 04:35:31 #action tpatil backport https://review.openstack.org/#/c/437312/ to stable/ocata 04:35:45 tpatil: I hv just put that action to you. 04:36:21 samP: ok 04:36:27 tpatil: thank you 04:37:31 samP: we need functional test soon to find out such issues 04:38:49 tpatil: agree, lets discuss pike work items including functional tests. 04:39:00 samP: sure 04:39:56 tpatil: sorry that I didn't have time to sort the things. Lets have that discussion on next week IRC, I'll create a etherpad for this. 04:40:17 any other bus to discuss? 04:40:22 samP: Ok 04:40:44 samP: ok 04:41:47 Is this still valid https://bugs.launchpad.net/masakari-monitors/+bug/1661517 ? 04:41:47 Launchpad bug 1661517 in masakari-monitors "masakari-instancemonitor fails to start with error “required config option auth-url”" [Undecided,New] - Assigned to Pooja Jadhav (poojajadhav) 04:42:55 samP: I will request Pooja to try it on the latest masakari-monitors code 04:43:06 tpatil: sure, thanks 04:43:30 if no other bugs, lets jump into discussion 04:43:42 #topic discussion points 04:44:36 As I said previously, I will create a etherpad for pike work items. Please add any item you would like to work or any proposals are welcome. 04:45:09 samP: Ok 04:45:09 I would like to discuss those items and prioratize them in our next IRC. 04:45:27 samP: OK 04:47:10 #link https://etherpad.openstack.org/p/masakari-pike-workitems 04:47:52 Here ^^ is the link for Pike work items 04:48:52 as of now, it has no contents :) 04:49:10 I will also create wiki item for pike release schedule which will be same as other official projects 04:49:53 sagara: Just created it :), I will populate it with my thoughts later.. 04:50:11 I will add " Implement auto_priority and rh_priority recovery_methods" as pike work item. Dinesh has already pushed patch for this implementation 04:50:24 #link : https://review.openstack.org/#/c/433669 04:50:39 tpatil: thanks, that will do 04:51:28 any other points to discuss? 04:52:00 ah.. I will share PTG updates for masakari in mail. 04:52:53 if no other topics, lets jump in to AOB 04:52:58 #topic AOB 04:57:23 Nothing else to discuss, lets finish the meeting.. 04:57:54 Thank you all for your time... 04:57:54 samP: yes 04:58:02 samP: I heard from you, TC are willing to add mentor to our project. When does it starts?, do we need anything to prepare? 04:59:09 sagara: I have reach them first. I will post that details in mail to you all. 04:59:40 sagara: no need to specially prepare 04:59:53 samP: thank you 05:00:02 OK, then its almost time, lest finish 05:00:11 thank you all... 05:00:38 Thank you. 05:00:39 #endmeeting