if the disk is under heavy pressure the default 5s seem a little bit optimistic to read the rs.info. What is the max value that makes sense resp. are there other options avail ?
Readinessprobe:
Tries to list all the databases. If you have like xxx’xxx databases, this call doesn’t make sense at all. Is there a way to change that resp. can a different readinessprobe applied ?
=> in both cases if there are too many failures the pods are getting shutdown. Clients can accept a long response time or even a timeout. But if they have to wait until the replicaset has done the restarted - it is just not a sensible way of doing the probes (imho) - or at least there should be other options.
Right now it is not possible to tune the probes through the operator.
Tries to list all the databases. If you have like xxx’xxx databases, this call doesn’t make sense at all. Is there a way to change that resp. can a different readinessprobe applied ?
To change the liveness and readiness probe. In kubernetes usually you could create your own probes. So is it not possible because the Percona Operator does not provide an interface or is it the MongoDB Operator ?