Discuss / brainstorm approach to Monitoring and alerting about apollo HOT 6 CLOSED

For Consul based alerting @armon is a fan of https://github.com/AcalephStorage/consul-alerts. Atlas will have an integration is PagerDuty shortly.

from apollo.

I really like prometheus. It has this for consul https://github.com/prometheus/consul_exporter

from apollo.

For microservices-infrastructure we want to decorate all logs and metrics with contextual metadata e.g., container_datacenter, container_host, container_name/id, container_version.

This should make it much simpler for app operators to diagnose issues across a multi-dc shared container runtime.

I'm interested in hearing your thoughts ... :-)

from apollo.

@wallies kk -- are you guys using Prometheus in production? Is it scaling well and easy to operate? I'm pretty lazy. ;-)

I'll spend some time hands on with it this week.

from apollo.

@keithchambers we arent using it in Production. I think it is based on Influxdb, which can scale really well for different datacenters and zones if you start to shard.

from apollo.

Closing this as we started to intregate prometheus via #526

from apollo.