04:00:22 <samP> #startmeeting masakari
04:00:23 <openstack> Meeting started Tue Aug 15 04:00:22 2017 UTC and is due to finish in 60 minutes.  The chair is samP. Information about MeetBot at http://wiki.debian.org/MeetBot.
04:00:24 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
04:00:26 <openstack> The meeting name has been set to 'masakari'
04:00:32 <samP> hi all../
04:00:35 <rkmrHonjo> hi
04:01:25 <samP> some people are not availbel due to summer vacation in Japan.
04:01:41 <tpatil> Hi
04:01:46 <samP> ah..
04:02:01 <samP> tpatil: you are here..
04:02:12 <tpatil> Even in India, today is a national holiday on occasion of Independence Day
04:02:28 <samP> tpatil: yes...\o/
04:02:31 <Dinesh_Bhor_> Hi all
04:02:37 <samP> Dinesh_Bhor_: hi
04:03:25 <samP> OK, then lets start
04:03:40 <samP> First, masakari release
04:04:32 <samP> every thing seems to ok, except rkmrHonjo and Dinesh_Bhor_ 7s patch
04:04:47 <samP> s/7/'/
04:05:08 <samP> #link https://review.openstack.org/#/c/493534/
04:06:11 <tpatil> The config option name should be 'ignore_error_instances' as mentioned in rkmrHonjo patch
04:06:13 <samP> rkmrHonjo: are you ok to abandon your's in favor to ^^
04:06:24 <tpatil> #link: https://review.openstack.org/#/c/493476/
04:06:48 <rkmrHonjo> samP: yes, I can abandon my patch. I don't stick to mine.
04:07:00 <rkmrHonjo> Should I abandon now?
04:08:10 <samP> rkmrHonjo: sure, in that case could you review Dinesh_Bhor_ patch and help him
04:08:30 <samP> tpatil: agree.
04:09:00 <tpatil> There are few minor review comments on Dinesh's patch which needs to be fixed, I will also give one comment to rename the config option from 'evacuate_error_instances' to 'ignore_error_instances'
04:09:16 <rkmrHonjo> samP: OK. I've abandoned my patch. I'll review Dinesh's patch.
04:09:22 <samP> tpatil: sure, thanks
04:09:44 <samP> Dinesh_Bhor_: can you fix it, or you are in vacation?
04:10:43 <samP> In order to release, I have a minor dependency #link https://review.openstack.org/#/c/493434/
04:10:56 <Dinesh_Bhor_> samP: I am. I can do the changes but I don't have multinode environment on my laptop to test it after changes
04:12:14 <tpatil> samP: Can we hold on to cut branch until tomorrow evening?
04:12:51 <samP> tpatil: yes..why?
04:13:30 <tpatil> samP: So that Dinesh can test the changes before pushing it for review
04:13:54 <samP> tpatil: sure.
04:14:38 <Dinesh_Bhor_> samP: tpatil: OK, thank you
04:14:46 <samP> No strik date for release, since we do not follow the official release cycle. However, beeing late is not that good
04:15:02 <samP> tpatil: tomorrow is OK
04:15:30 <samP> Let's do the release for all 3 projects tomorrow.
04:16:23 <tpatil> samP: Sure
04:16:35 <rkmrHonjo> samP: agree.
04:16:36 <samP> Dinesh_Bhor_: tpatil: rkmrHonjo: Thank you..
04:16:59 <samP> OK, then let't move to next item..bugs
04:17:07 <samP> #topic bugs
04:17:30 <samP> #link https://bugs.launchpad.net/masakari/+bug/1710570
04:17:31 <openstack> Launchpad bug 1710570 in masakari "Flow 'instance_evacuate_engine': InvalidInput: Invalid input received: Expecting to find domain in project" [Undecided,In progress] - Assigned to Liyingjun (liyingjun)
04:17:55 <samP> patch was there, but seems need to fix the unit tests
04:18:04 <Dinesh_Bhor_> The reporter discussed this issue with me yesterday on masakari channel
04:18:17 <samP> Dinesh_Bhor_: thanks
04:19:00 <Dinesh_Bhor_> and it is reproducible on environment where keystone is not deployed on uwsgi, because I don't see this issue on my environment where keystone is on uwsgi be default with devstack
04:19:16 <Dinesh_Bhor_> s/be/by
04:20:30 <samP> Dinesh_Bhor_: thanks for the info.
04:20:35 <tpatil> I don't think it's an issue of uwsgi, maybe how keystone is configured
04:21:19 <samP> well, If we do this #link https://review.openstack.org/#/c/493711/1/masakari/compute/nova.py
04:21:38 <samP> is it possible to use same options for V<3 and V3?
04:22:28 <samP> I need to review this. I will do it later..
04:22:29 <Dinesh_Bhor_> samP: Yes, I think so
04:22:56 <samP> Dinesh_Bhor_: thanks.. I will check
04:23:13 <samP> any other bugs?
04:24:14 <samP> if not, lets move to discussion, otherwise please bring them to AOB
04:24:24 <samP> #topic Discussion points
04:24:53 <samP> Sorry, application for oficial project is half done. I will finish it soon.
04:25:15 <samP> (2) Installation docs
04:25:29 <samP> we have alreday discuss this in past. any new issues?
04:25:59 <rkmrHonjo> no.
04:26:06 <samP> ok then,
04:26:29 <samP> (3) Should the 'error' instances be only stopped after recovery or stop and reset to 'error' again?
04:27:02 <samP> I answered this, any further issues?
04:27:10 <tpatil> samP: #3 is already addressed and the patch is merged
04:27:20 <samP> tpatil: yes. thanks.
04:27:32 <samP> (4) recovery method customization
04:28:11 <samP> I think we can discuss about this with Mistaral team at PTG
04:28:24 <tpatil> samP: Abhishek has pushed updated specs for review
04:28:46 <samP> tpatil: thanks, I will review
04:28:50 <tpatil> samP: we have no dependency on mistral except restart service issue after actions are registered
04:29:35 <samP> tpatil: ah..yes.. I am conficed with register action on the fly
04:30:07 <tpatil> samP: we can discuss about this issue with Mistral team in PTG
04:30:10 <samP> s/conficed/confused
04:30:46 <tpatil> Dinesh_Bhor: Please  let us know if there is anything else required in Mistral
04:30:48 <samP> tpatil: sure, do you think its better to put this in Mistral team agenda?
04:31:10 <Dinesh_Bhor_> tpatil: No other changes, just the restart issue
04:31:54 <tpatil> samP: Sure, I will add this point in the etherpad
04:32:03 <samP> tpatil: thanks.
04:32:05 <samP> Dinesh_Bhor_: thanks
04:32:36 <samP> (5) How do we find hosts if we forget the relational of segments-hosts?
04:32:47 <rkmrHonjo> I heard this opinion from a masakari user.
04:32:56 <rkmrHonjo> He said that "It will be hard to find specific hosts if operator forgets the relational of segments-hosts."
04:32:57 <samP> rkmrHonjo: thanks.
04:33:11 <rkmrHonjo> What do you think about it?
04:34:11 <rkmrHonjo> He can solve this issue by making scripts, but I think that it is not kindly.
04:34:44 <tpatil> What about get host API?
04:35:22 <tpatil> We need to pass segment id,
04:35:41 <rkmrHonjo> tpatil: For example, he want to update host. But, he forget to relational of segments-hosts. So he can't specify correct segment-id.
04:36:19 <rkmrHonjo> In current masakari, he should call host list api for each segments.
04:36:54 <rkmrHonjo> s/forget/forgot/
04:37:37 <tpatil> one solution is to add a new command in masakari client to return failover segment for a given host uuid or name
04:38:43 <rkmrHonjo> tpatil: Yes. Your idea will use only existing APIs, doesn't it?(Not create new API)
04:39:00 <rkmrHonjo> s/doesn't/won't/
04:39:02 <samP> tpatil: yes, that is rkmrHonjo's 2nd option
04:39:02 <tpatil> rkmrHonjo: Yes
04:39:18 <samP> do we need to extend APIs for this?
04:40:21 <tpatil> samP: it's possible
04:40:23 <samP> if some one wants to know the details about a host, he must know the which segment the host is.
04:40:36 <samP> ^^in current masakari
04:41:37 <samP> /host/{segment-id}/list
04:41:41 <Dinesh_Bhor_> samP: yes you are right as the hosts API "os-hosts" is dependent API
04:42:27 <samP> so, what rkmrHonjo is asking, some thing like /host/list
04:42:55 <samP> which shows the host name, segment, current status, ...etc..
04:44:16 <rkmrHonjo> samP: I afraid that making new API is too much. But we should make new API if many users want to this...
04:45:45 <rkmrHonjo> (I couldn't decide yet. sorry...)
04:46:10 <samP> rkmrHonjo: can you check waht need to be changed for (1) new API (2) do it in masakariclient ?
04:46:38 <samP> rkmrHonjo: let's discuss more in next meeting.
04:47:00 <rkmrHonjo> samP: OK. I'll report it in next meeting. And I'll decide my opinion.
04:47:33 <samP> rkmrHonjo: thanks. based on your details, lets discuss which is the betst way to proceed
04:47:58 <samP> (6) Pike work items
04:49:07 <samP> Most of them are under discussion or not implemented (my parts..)
04:49:12 <samP> sorry
04:49:38 <samP> Lets move them to Q..
04:50:14 <tpatil> samP: Are you planning to create an ether pad for Queens before PTG?
04:50:24 <samP> In next meeting, lets discuss what left fro Queens
04:50:30 <samP> tpatil: yes.
04:50:41 <samP> I will send a mail to ML with the link
04:50:54 <tpatil> samP: Ok, Thanks
04:51:10 <samP> (7) "masakari-manage db purge": support for purging notification records
04:52:01 <samP> I think spec merged, and patchs are in the review.
04:52:21 <tpatil> samP: yes
04:52:23 <samP> Lets merge this for Queens
04:52:42 <tpatil> samP: Sure
04:52:47 <samP> OK, then aob
04:52:52 <samP> #topic AOB
04:53:14 <rkmrHonjo> Can you execute notification API in latest devstack?
04:53:24 <samP> Masakari monitors old shell scripts will be removed in queens
04:53:33 <rkmrHonjo> oops, sorry.
04:54:00 <samP> rkmrHonjo: its oK, "latest"=current masater?
04:54:17 <rkmrHonjo> yes.
04:54:22 <rkmrHonjo> I built my environment by devstack yesterday. And I called notification API. As a result, masakari-engine raised below error when masakari-engine called nova API.
04:54:31 <rkmrHonjo> 2017-08-14 17:54:12.597 18514 WARNING keystoneauth.identity.generic.base [req-8e3807d0-6aca-494c-a814-7019895e6d37 nova - - - -] Failed to discover available identity versions when contacting http://127.0.0.1/identity. Attempting to parse version from URL.: ServiceUnavailable: Service Unavailable (HTTP 503)
04:54:50 <rkmrHonjo> (I used devstack-masakari-plugin.)
04:56:23 <samP> I am currently not using devstack-masakari-plugin, however no such isuue. But my devstack is 2 weeks old.
04:56:24 <rkmrHonjo> I don't try other environment yet. There is a possibility that this happen in only my environment.
04:56:49 <tpatil> rkmrHonjo: can you please report this issue in LP?  Dinesh will check this problem tomorrow at his end
04:56:57 <samP> rkmrHonjo: I will try devstack-masakari-plugin
04:57:04 <samP> tpatil: thanks.
04:57:15 <rkmrHonjo> taptil: OK. And, I try to re-build environment on other machine.
04:57:59 <samP> One other thing; Migrate from launchpad to Storyboard
04:58:34 <samP> All the openstack projects are moving away from launchpad to storyboard and we should too
04:58:45 <samP> there are no dead line for this.
04:59:11 <samP> I will check this with #storyboard
04:59:17 <samP> We are out of time
04:59:32 <samP> please use #openstack-masakari or ML for further discussion
04:59:38 <samP> Thank you all
04:59:44 <samP> #endmeeting