13:00:45 <witek> #startmeeting monasca
13:00:46 <openstack> Meeting started Tue Nov 26 13:00:45 2019 UTC and is due to finish in 60 minutes.  The chair is witek. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:00:47 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:00:49 <openstack> The meeting name has been set to 'monasca'
13:01:22 <witek> hello everyone
13:01:44 <chaconpiza> Hi
13:01:48 <dougsz> hey all
13:02:01 <witek> hi, nice to see you
13:02:13 <witek> light agenda for today:
13:02:16 <witek> https://etherpad.openstack.org/p/monasca-team-meeting-agenda
13:02:37 <witek> #topic InfluxDB proxy
13:03:12 <witek> in the recent email thread at openstack-discuss someone has pointed to influxdb-proxy tool
13:03:19 <witek> https://github.com/shell909090/influx-proxy
13:03:45 <witek> it extends the concept of InfluxDB relay
13:04:15 <dougsz> hmm, yes, I think I still prefer the Kafka consumer groups thing for HA
13:04:31 <witek> and from reading the README, it provides both replication and sharding
13:04:50 <witek> dougsz: agree, I wanted to point to the concepts implemented
13:05:07 <dougsz> One less service and that project doesn't look particularly dependable
13:05:28 <dougsz> makes sense, thanks witek
13:06:13 <witek> yes, what I wanted to point is that they hold the keymap and redirect the messages and queries to the particular InfluxDB instance
13:06:45 <witek> we could implement it much nicer in monasca-api
13:06:50 <dougsz> Solid idea - a bit like a placement group in Ceph perhaps
13:08:03 <dougsz> It would be nice to try TimescaleDB at the same time to see if it lives up to the hype
13:08:03 <witek> I have added the repository URL to our Story:
13:08:21 <witek> https://storyboard.openstack.org/#!/story/2005620
13:09:35 <witek> yes, we could try TimescaleDB as well
13:10:48 <witek> some more comments on this topic?
13:11:12 <dougsz> I guess it could also be used for mapping tenants to dedicated InfluxDB instances
13:11:32 <Dobroslaw> looks like influx-proxy is using redis
13:12:58 <witek> dougsz: yes, sharding on tenants would be the simplest logic
13:14:51 <witek> Dobroslaw: do we need it? I think we could store the keymap in mysql
13:15:31 <dougsz> yeah - that would make more sense
13:15:33 <Dobroslaw> just pointing that influx-proxy would need 4th db in monasca
13:15:59 <Dobroslaw> good idea with keymap in mysql
13:16:06 <witek> but I wasn't talking about using it, just showing the concept
13:16:11 <Dobroslaw> a, ok
13:17:15 <witek> #topic Reviews
13:17:33 <witek> no much progress here since last week
13:17:45 <witek> I've updated the board:
13:17:54 <witek> https://storyboard.openstack.org/#!/board/190
13:18:24 <witek> we've got two stories needing review
13:18:33 <witek> and one needing help
13:19:08 <witek> devstack plugin for merged api has open comments
13:19:41 <witek> apart from these there is also:
13:19:41 <witek> https://review.opendev.org/674204
13:19:54 <witek> which unlocks events-api repo
13:21:01 <witek> any other reviews we should look at?
13:22:06 <witek> any progress on ELK upgrade?
13:23:18 <witek> #topic AOB
13:23:21 <Dobroslaw> Arseni have some problems with updating Kibana plugin
13:23:50 <witek> should we proceed without Kibana plugin first?
13:24:07 <arseni-lipinski> hi
13:24:10 <arseni-lipinski> you may proceed
13:24:34 <witek> are all changes up-to-date now?
13:25:10 <arseni-lipinski> without the plugin they should be
13:25:49 <witek> could you point me to the story please?
13:26:09 <arseni-lipinski> https://storyboard.openstack.org/#!/story/2006376
13:27:10 <witek> thanks, added to the Board
13:28:19 <witek> on the logs tempest-job front there were less failures in the last days
13:28:28 <witek> but the problem is not gone
13:29:01 <witek> http://zuul.openstack.org/builds?job_name=monascalog-python3-tempest
13:29:31 <witek> what we could try is to disable deploying Kibana in CI
13:29:43 <arseni-lipinski> about the plugin: the problem i'm working on now is that after creation of needed index-pattern kibana still won't see it until the page refresh. in old kibana page refresh could be achieved by replying with html, that contained page refresh command in <script>. unfortunately, new kibana throws an unexpected token error when trying to pass html to kibana. i'm out of ideas. something will maybe come up, but for now i don't see a way
13:29:43 <arseni-lipinski> to do so.
13:30:27 <witek> arseni-lipinski: that's sad
13:31:56 <arseni-lipinski> i'll finalize the Elkstack update in log-api, remembered there was an issue with log-metrics, i'll fix it asap
13:32:21 <witek> thanks
13:33:03 <witek> any other topics for today?
13:33:31 <dougsz> Nothing from me
13:34:06 <witek> let's try to make some progress on open reviews till next week
13:34:22 <witek> and also try to solve the tempest job issue
13:34:26 <witek> https://etherpad.openstack.org/p/monascalog-python3-tempest
13:34:43 <witek> thanks for joining
13:34:47 <witek> and see you next week
13:35:18 <witek> #endmeeting