Hi
We are using Aerospike cluster of 3 nodes with replication=2 for the Namespace.
Whenever we restart a single node in cluster the java client start throwing below exception:
com.aerospike.client.AerospikeException$Timeout: Client timeout: timeout=10000 iterations=1
failedNodes=0 failedConns=0
Ideally if my 2 nodes are up, it should not throw this exception, as replica copy is still there.
Thanks.
When you bring a server down, Aerospike doesn’t really do this gracefully. It’s the same as suddenly losing a server. Couple of questions to help you from here:
Are you experiencing issues even after the server goes down (server down time + 3 seconds)?
What type of call is throwing timeout? Read? Write? Query?
Yes Albot, we were experiencing the issue even after server went down ( i have restarted the server but it took 15 mins to join in the cluster, due to rebalancing the data)
We were getting the errors for write.
Rebalance doesn’t prevent a node from joining the cluster (in fact you cannot rebalance without being in a cluster). I believe you meant loading data from disk.
Right , Agree with you, it was loading the data