Kubernetes Monitoring

Semaphore-like monitoring following drill-down principle

Overview

dNation Kubernetes Monitoring allows us to see status of our infrastructure and applications at a glance using semaphore (green/orange/red) principle:

  • Green – All is fine, I can go to lunch
  • Orange – Something is wrong, but it’s not urgent, I can go to lunch but have to have a look afterwards
  • Red – something is seriously wrong, I can’t go to lunch right now

Basic design principles:

  1. Intuitive – Green, orange and red colors signaling whether or not your action is needed
  2. Drill – down – if you want details why something is green, orange or red, just click it
  3. Relevant information only – provide only metrics relevant for this particular area of interest and drill-down level, side-by-side with logs (experimental feature)

Monitoring targets

Basic use case

  • You received 17 alerts at 3 AM
  • You look at dashboard and see multiple problems, application A is in red, some PVCs are red, while k8s cluster itself is green; so you know where to start to look for root cause (non-green color)
  • You click PVC stat-panel, get a list of PVCs, all of them green, 1 of them red
  • It turns out that red PVC is used by application A
  • You fix PVC issue, and suddenly whole dashboard becomes green, so you have some indication that things are probably all right now
  • As a result, incident response time is improved

Commercial support

This project has been developed, maintained and used in production by professionals to simplify their day-to-day monitoring tasks and reduce incident reaction time.

Commercial support is available, including 24/7, please contact us.

AWS Marketplace

This project is awailable also at AWS Marketplace​.

Credits

Following projects are used:

User Reviews

Source code

Open Source Software, Apache 2.0 License.