6

I need to setup some jenkins jobs that probe the cluster every X minutes, and make sure everything is working and functional. Are there any good off-the-shelf blackbox probers for the Kubernetes API , either opensource or provided by some vendor.

Although I can write my own shell scripts to do test deployments and check things that everything is in good shape , But i am looking for blackbox test suite with some reasonable coverage like:

  • checking deployements are working
  • replcaSet are doing the replicas
  • service discovery and dNS are functional
  • configmaps and secrets can be created , and the secret encryption is working
  • apiServer latency is not too high
  • horizontal pod auto scaling is working
  • service can be created and user can connect from outside
Ijaz Ahmad
  • 11,198
  • 9
  • 53
  • 73
  • In what way is the existing prometheus machinery insufficient for this task? And, of course, "working and functional" means **a lot** of different things to different people. What risks have you seen that you would like to prevent; because the list isn't "everything that can go wrong," since that's almost infinite. – mdaniel Oct 28 '18 at 03:23
  • promethous does the is white box monitoring – Ijaz Ahmad Oct 28 '18 at 07:50
  • @MatthewLDaniel there is a need to understand the difference between white box monitoring and blackbox monitoring – Ijaz Ahmad Oct 28 '18 at 08:23
  • working and functional has only one meaning for a kubernetes component , it doenst have alot of meanings if I say the APIserver is accepting and processing requests , and the sheduler is schedling the pods , and Service discovery is operational – Ijaz Ahmad Oct 28 '18 at 08:30
  • Take a look at Heptio Sonobuoy for everything except the last one. – Keith Burdis Oct 29 '18 at 08:33

0 Answers0