Roadmap

This is a list documenting the work that is being considered;

Short term

• make it easier to determine the data retention policy in graphite from the ansible playbook. The caveat is that the lowest granularity is 10s - since this syncs with the collectd interval
• add an alerting dashboard : create a notification channel, and provide basic alerts from key metrics to this channel
• examine whether dashUpdater should be used to set up the notification channel for alerting through Grafana's http api
• update the osd collector to report on recovery and backfill latency per OSD, then review how/where these metrics can be used (at-a-glance and osd-latency dashboard to begin with)

Medium/Long term

• investigate tagging - can this be used to tag nic interfaces as public/cluster for frontend/backend charts?
• review a migration to prometheus, replacing graphite - but retaining grafana for visualisation
• review the benefit of including haproxy based statistics within the dashboard(s)
• review the integration options with cephmgr