| *** k_mouza has joined #openstack-monasca | 00:15 | |
| *** k_mouza has quit IRC | 00:20 | |
| *** adriancz has quit IRC | 00:22 | |
| *** k_mouza has joined #openstack-monasca | 01:31 | |
| *** k_mouza has quit IRC | 01:36 | |
| *** k_mouza has joined #openstack-monasca | 02:45 | |
| *** k_mouza has quit IRC | 02:50 | |
| *** k_mouza has joined #openstack-monasca | 03:27 | |
| *** k_mouza has quit IRC | 03:31 | |
| *** brtknr has quit IRC | 06:17 | |
| *** vishalmanchanda has joined #openstack-monasca | 06:39 | |
| *** nikparasyr has joined #openstack-monasca | 06:53 | |
| *** dougsz has joined #openstack-monasca | 07:32 | |
| *** tosky has joined #openstack-monasca | 07:39 | |
| *** brtknr has joined #openstack-monasca | 08:47 | |
| *** witek has joined #openstack-monasca | 09:07 | |
| *** witek has quit IRC | 09:09 | |
| *** k_mouza has joined #openstack-monasca | 09:29 | |
| *** witek has joined #openstack-monasca | 10:39 | |
| nikparasyr | hello :). a couple of notes about the influx HA setup. couple of weeks ago i mentioned setting the legacy_kafka_client_enabled = false on persister.conf, allowed it to work properly. last week i changed that option on api.conf as well but that lead to errors -- im not sure on which service now and logs are flushed. I saw that upstream the default value is set to false. So my remark is that we might need to check if it works pr | 12:23 |
|---|---|---|
| nikparasyr | operly if set on api.conf | 12:23 |
| nikparasyr | im currently running monasca train on a test cluster, so i can enable it again and provide logs etc. let me know | 12:24 |
| nikparasyr | apologies for not setting it and mentioning it earlier | 12:24 |
| nikparasyr | another thing, which i sort of want to see if there are ppl interested in this. This is mostly a kolla-ansible "proposal" but i want to run it through here. We have deployed a stand-alone 3-node monasca cluster with external keystone with kolla-ansible. this works quite well. We have 2 configurations, one for the openstack cluster and one for the monasca cluster. The issue we have now is that with k-a we cannot config the mona | 12:28 |
| nikparasyr | sca role to only deploy the agent on the openstack nodes. | 12:28 |
| nikparasyr | I am thinking of making functionality on the k-a monasca role to define external_monasca and then only services such as the agent are configured and deployed on the nodes. similarly ceilometer config is changed to make it work for monasca etc | 12:29 |
| nikparasyr | Is this something that makes sense and we expect more ppl to need/require it? if yes i could make a blueprint on k-a about this change and start working on it (not in the next couple of weeks though) | 12:30 |
| dougsz | nikparasyr: good questions - will reply in an hour or so when I am out of the current meeting. | 13:03 |
| *** piotrowskim has joined #openstack-monasca | 13:46 | |
| dougsz | nikparasyr: So about the first point - I'm working on that now. It seems to work fine on master, but we're going to run Monasca Tempest against it in CI to be sure | 14:13 |
| dougsz | About the other question - it's a good idea to decouple Monasca Agent from the Monasca to support what you need. | 14:15 |
| dougsz | We already have a similar mechanism with Prometheus + Prometheus Exporters threre | 14:15 |
| dougsz | It would get my +2! | 14:15 |
| dougsz | I wouldn't worry about a blueprint for it, just propose it when you have time. | 14:16 |
| dougsz | One thing I was thinking about the consumer groups change is that we could use sticky sessions so the user hits the same InfluxDB instance each time they refresh | 14:18 |
| dougsz | And also for Tempest | 14:18 |
| dougsz | which suffers race conditions with multiple Influx instances | 14:19 |
| nikparasyr | ok. i think the prometheus + prometheus exporters was introduced recently so i will have a look into how you decoupled that one | 14:20 |
| nikparasyr | i might need some help with some "design" decisions. also i've only made one upstream change to k-a (yday) so it might take some time to get used to the workflow and conventions used. but only one way to learn | 14:21 |
| nikparasyr | about the sticky sessions i've seen the story/todo on the WIP you have. i havent worked a lot with sticky sessions on haproxy but it seems like a good addition | 14:21 |
| nikparasyr | another thing which i am a bit sceptical is that on a 3-node setup you get 3 grafana's which is... well weird. for the time being i changed the inventory to only deploy one | 14:24 |
| dougsz | nikparasyr: no problem - I think you have got the workflow correct and happy to help out with any design | 14:41 |
| dougsz | Are you accessing Grafana via the VIP/HAProxy? | 14:42 |
| nikparasyr | yep | 14:42 |
| dougsz | The idea is that all Grafana instances store state in the MariaDB cluster | 14:42 |
| nikparasyr | hmmm, wasnt aware of that. so basically configuring lets say a data source to one will make the datasource to all 3, etc with dashboards | 14:43 |
| dougsz | So to the end user won't be aware of which instance they use since they are all identical | 14:43 |
| dougsz | yeah, that's correct, because they are configured to use the DB | 14:43 |
| nikparasyr | oh wasnt aware of that. that's great actually :) thanks! | 14:44 |
| dougsz | np! For loading dashboards this is one option: https://github.com/stackhpc/ansible-role-grafana-conf | 14:44 |
| dougsz | We were planning on adding to KA, but no time yet | 14:44 |
| dougsz | And this for generating dashboards: https://github.com/stackhpc/grafana-reference-dashboards | 14:45 |
| *** irclogbot_0 has quit IRC | 14:52 | |
| *** irclogbot_2 has joined #openstack-monasca | 14:59 | |
| nikparasyr | ok thanks for the dashboards as well :) | 15:06 |
| *** nikparasyr has left #openstack-monasca | 15:16 | |
| *** k_mouza has quit IRC | 16:09 | |
| *** k_mouza has joined #openstack-monasca | 16:28 | |
| *** dougsz has quit IRC | 16:31 | |
| *** k_mouza has quit IRC | 16:33 | |
| *** witek has quit IRC | 16:45 | |
| *** k_mouza has joined #openstack-monasca | 16:46 | |
| *** k_mouza has quit IRC | 16:50 | |
| *** k_mouza has joined #openstack-monasca | 16:57 | |
| *** k_mouza has quit IRC | 17:01 | |
| *** k_mouza has joined #openstack-monasca | 17:15 | |
| *** k_mouza has quit IRC | 17:19 | |
| *** k_mouza has joined #openstack-monasca | 17:35 | |
| *** k_mouza has quit IRC | 17:39 | |
| *** k_mouza has joined #openstack-monasca | 17:54 | |
| *** k_mouza has quit IRC | 17:58 | |
| *** k_mouza has joined #openstack-monasca | 18:10 | |
| *** k_mouza has quit IRC | 18:14 | |
| *** vishalmanchanda has quit IRC | 18:29 | |
| *** k_mouza has joined #openstack-monasca | 18:31 | |
| *** k_mouza has quit IRC | 18:39 | |
| *** k_mouza has joined #openstack-monasca | 19:22 | |
| *** k_mouza has quit IRC | 19:27 | |
| *** k_mouza has joined #openstack-monasca | 19:33 | |
| *** k_mouza has quit IRC | 19:37 | |
| *** k_mouza has joined #openstack-monasca | 19:42 | |
| *** k_mouza has quit IRC | 19:47 | |
| *** k_mouza has joined #openstack-monasca | 20:05 | |
| *** k_mouza has quit IRC | 20:09 | |
| *** piotrowskim has quit IRC | 23:12 | |
| *** tosky has quit IRC | 23:21 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!