Clock skew across a Cassandra cluster

Clock skew across a Cassandra cluster can cause synchronization and replication issues.

For example, data might reappear after deletion, or seemingly successful modifications might not be visible.

Remedy

Procedure

Use the Network Time Protocol daemon to synchronize clocks for all nodes in the cluster.

Note: Acceptable skew is 10 milliseconds.

For more information about how to install the Network Time Protocol daemon, see your operating system's documentation.