Thursday, 2019-02-14

*** akhil_jain has joined #openstack-watcher05:11
*** akhil_jain has left #openstack-watcher05:11
*** licanwei has joined #openstack-watcher07:44
*** chenke has joined #openstack-watcher07:46
licanweijosecastroleon: hi~07:46
josecastroleonsorry i went for a coffee07:53
josecastroleonwe did an investigation during summer about Watcher, and we tried several strategies07:55
licanweijosecastroleon: so what's your plan?07:56
josecastroleonwe found some rough edges but in general we were pretty happy with the tool07:56
josecastroleonwe have some use cases that we want to contribute that are variations of the existing ones07:57
josecastroleonwe are using extensively cells for scale the infrastructure out, and in some cases they are not well balanced07:57
josecastroleonwe can use the VM workload balancing strategy but we need to limit it somehow to a subset of hosts07:58
licanweiabout your use cases, can you give me some details ?07:58
josecastroleon1) VM workload balancing for cells07:59
licanweiwe can set audit scope through parameter07:59
josecastroleon2) Scheduler for preemptible instances07:59
licanweiwe have a scope parameter in audit08:00
josecastroleon3) continuous audit on hyperconverged servers08:00
josecastroleonon the scheduler of preemptibles, normally we receive hardware and it takes some time to be filled by our clients. As we have other workloads that can get benefit, we can set a continuous audit looking for holes and spawning new instances there08:01
josecastroleonon the continuous audit for hyperconverged, we have a setup in which the compute and storage lives together. In this setup we didn't have any fixed limit to protect CPU or IO and we want to do again workload balance keeping some free resources for IO08:03
josecastroleonthose are the use cases that we want to contribute at the moment08:04
josecastroleonAs rough edges, and linked to what I said previously, when fetching data on a large cloud, some operations on nova are pretty expensive.08:05
josecastroleonretrieving data from 10k servers and all the machines hosted takes too long08:05
josecastroleonthis is why we would like to add a scope08:05
licanweiI got it08:07
licanweiyou want to add a scope to limit Watcher datamodel08:08
josecastroleonfinally, we are not using neither monasca, nor ceilometer nor gnocchi for the metrics. We have all the metrics in a set of influxDBs with grafana in front.08:09
licanweigood idea :)08:09
josecastroleonso we may add a datasource for grafana-proxy, if someone is interested in it08:10
licanweiHow you get metrics from influx? Rest API or sql?08:11
josecastroleonour monitoring team has almost a influxdb instance per metric, so we are using grafana-proxy for that (rest api)08:13
licanweiif you don't use ceilometer, how you get metrics from data sources?08:17
licanweiand about rough edges, Watcher need add a new scope for datamodel and a new datasource for metrics, right?08:22
josecastroleonI need to run now for a meeting, i hope to come back soon08:23
licanweiwe can discuss tomorrow if no time today :)08:24
*** licanwei has quit IRC09:24
*** chenke has quit IRC09:25
*** josecastroleon has quit IRC15:01
*** ianychoi has joined #openstack-watcher17:06
*** Dantalion has joined #openstack-watcher17:58

Generated by 2.15.3 by Marius Gedminas - find it at!