Sensu as monitoring system

[altvnk]

We could use sensu as a monitoring system for core services. Also it can be used for metric shipping.

[corebug]

Sensu is great, but, on the other hand, can be very tricky to use since it's a development framework and not a "monitoring system" . Anyway, i'd vote for Sensu, yes.

[altvnk]

Obviously it's more like an "monitoring router", maybe "monitoring framework", but exactly not "development"
All you need is to have a check script that returns standard exit codes, nothing tricky i think.

[kenjones-cisco]

Is the idea to leverage sensu instead of https://github.com/kubernetes/kubernetes/tree/master/cluster/addons/cluster-monitoring which seems to be based on cAdvisor?

[altvnk]

Basically idea was not only monitor but also do some health checks and alerts based on that. And we can continue using cAdvisor to get metrics and other stuff.

17 авг. 2015 г., в 18:37, Kenny Jones [email protected] написал(а):

Is the idea to leverage sensu instead of https://github.com/kubernetes/kubernetes/tree/master/cluster/addons/cluster-monitoring which seems to be based on cAdvisor?

—
Reply to this email directly or view it on GitHub.

[kenjones-cisco]

Sounds good.

[corebug]

Cadvisor can be used as a data source for, I guess, any metrics collection system. And Sensu, indeed, is great in terms of healthchecks and some alerts in reaction to those.