Monitoring docker in production

[deedubs]

• Deploy a cAdvisor container and scrape it via prometheus.
• Set alerts for common fatal conditions
• Demonstrate application specific alerts
• Expose the prometheus metrics via a Grafana Dashboard.

[hairyhenderson]

@deedubs Solid! Would you be willing to deliver the talk?

[deedubs]

Yeah for sure!

[hairyhenderson]

@deedubs - do you want to schedule this talk in for the October meetup? (October 26)

[deedubs]

Hey. I'm spinning up in a new role so don't have the bandwidth ATM.

[hairyhenderson]

Ok @deedubs - no worries. We'll look at scheduling this another time!

[ffoysal]

@deedubs Just wondering is it happening soon !!!

[deedubs]

@ffoysal I'm scheduled to do a similar talk at the k8s ottawa meetup. Someone else can take up the reigns of this topic if they don't want the k8s context or want it at the docker meetup.

[hairyhenderson]

I could do this talk - been doing lots of tinkering with prometheus+grafana lately... But it doesn't have to be me 😉

Some further ideas too:

• incorporating the Docker Engine's experimental prometheus endpoint
• deploying Grafana HA with a multi-node MySQL/MariaDB back-end