Hi,

Also note that we are integrating grafyaml to the config repo so that
grafana dashboards can be proposed and updated by regular user too.

We also plan to use grafyaml for dashboards configuration.

Here are some more thoughts:

Dependending on how the metrics are pushed, we may need some kind of
authorization mechanism and a job secret to allow external clients to
push new metrics.

Yes, I thought about it too. In case of rdo cloud tenant jobs we can just filter by IP,

but in case we need to send from upstream clouds to server in RDO SF seems like it

should be authorization. But I'm not sure how it could be done from openstack-infra clouds.

Job secret in repo is not a secret..

Openstack-infra graphite is firewalled too and doesn't allow sending data to it from jobs

right now as well. Another option is pulling data from jobs logs, but it's very complicated

and not preferable.

It seems like we could setup post run to push job metrics. Perhaps
we could leverage ara sqldump to extract per task duration.

Software Factory may also automatically setup job duration graph
dashboard per project, here is a new user-story to track this work:
https://tree.taiga.io/project/morucci-software-factory/us/897

We have ansible module that parses ARA tasks output and convert it to

graphite data and sends to graphite server. It's actually how we do it right now.
It's easy to write the same for influxdb too.
https://github.com/openstack/tripleo-quickstart-extras/blob/master/roles/collect-logs/library/ara_graphite.py

Of course it should move to ARA in future.

Thanks

Best regards

Sagi Shnaidman