Cassandra repair mechanisms
Since data is partitioned and replicated onto multiple nodes, there is a possibility that replicas might be out of sync. This is usually due to slow responsiveness of certain nodes or nodes being down. If multiple nodes are out of sync, there is a possibility of reading stale data. There is also a certain case when deleted data reappears. To fix these issues, Cassandra provides three main repair mechanisms. These mechanisms ensure that the data is in sync most of the time. The repair mechanisms are as follows:
- Hinted handoff
- Read repair
- Anti-entropy repair
Hinted handoff
Whenever a node becomes unresponsive or down for a certain period of time, all the write requests to that node will fail. Cassandra keeps a copy of the partitions that were supposed to be written to the unresponsive node within a local hints table. The hints are stored for a certain configurable period of time. When the downed node comes up and starts gossiping with the rest of the nodes in the cluster...
