Monitoring demo (Prometheus)

Description

Monitoring and feedback is a critical piece of the software development lifecycle, and monitoring production systems and applications is key to a complete DevOps toolchain. Let's create a demo that illustrates our vision for how monitoring can be part of the GitLab experience.

Proposal

  • Create an MVP design for #23841 (closed) before Ben joins (Jan 9th)
  • Create an aspirational demo of monitoring, ideally before Ben joins, but at least before Mexico summit

Possibilities:

  • Latency, throughput, error rate
  • CPU load, memory
  • Other system monitoring (disk space, IO, ...)
  • Framework-specific APM (e.g. performance of individual endpoints, inspect rails stack for specific requests, etc.)
  • End-user monitoring (e.g. latency measured at the browser)
  • Business metrics/analytics
  • User metrics/analytics (e.g. clicks, flows, funnels, a/b testing)
  • Rollout / blue-green / canary monitoring (related to Deployboard https://gitlab.com/gitlab-org/gitlab-ce/issues/21413)
  • Feature monitoring (#24254 (closed))
  • Alerting

Links