Summary

Cassandra is a highly available, fault-tolerant, distributed database. However, sometimes data can get corrupted due to client application faults and other reasons. To handle such situations, Cassandra provides tools to back up and restore data to the last known state. Using the nodetool snapshot command, we can manually take snapshots of the data of a node. Restoring the snapshot from a node might not be consistents but restoring from all nodes' data will eventually become consistent. An incremental backup configuration allows the taking of automatic snapshots of node data. While loading SSTables in the bulk sstableloader is a great utility, using the sstableloader SSTables can be uploaded to different clusters with different ranges and replication factors too. The sstable2json command converts a SSTable to JSON format, which then can be converted back to SSTable using the json2sstable command and be loaded into a Cassandra node. The CQLSSTableWriter class APIs can be used to create...