Server timeout during read query at consistency LOCAL_ONE

There are several possible issues.

  • It may be a problem with a large amount of tombstones in cassandra. In this case we would suggest to either increase ‘tombstone_failure_threshold’ in cassandra.yaml or clean up tombstones.
  • There are not enough alive replicas. We would propose to increase the replication factor.
  • In some cases there is not enough request timeout. We are proposing to try to increase ‘read_request_timeout_in_ms’ in cassandra.yaml, if it is possible.