Stargate becomes non responsive

K8ssandra help
1

I have a default k8ssandra on GKE for my MemeGen demo. After a period of time the stargate pod stops responding. My solution is to delete it and then it recreates and works again.

kubectl delete pod k8ssandra-dc1-stargate-6f4c458d47-jhmc6

Some Questions:

  1. What would be the cause of this?
  2. What are next steps to trouble shoot?
  3. What can you guys recommend to make it more stable?
1 Like
2

The stargate pod stops responding on all interfaces (i.e. cql, rest, grapql) or just some of them? How many stargate pods do you have deployed and are they all experiencing the same issue or just one? Do you happen to have any logs or client side errors when the issue occurs?

  1. Hard to say without logs from stargate and/or Cassandra
  2. Confirming direct access to Cassandra is still working and pulling logs when the issue occurs
  3. Will depend on the cause of the issue but I’d suggest running at least two stargate pods if you aren’t already.
3

Single cass/stargate demo cluster. Default configs. I believe all interfaces fail, only see the app break because api doesnt respond to auth request. Will update here again with logs when the state fails again. FWIR: it takes about a week to break down.