Cassandra is optimized for writing large amounts of data very quickly. It writes all incoming data in an append-only manner in internal files called SSTables.
Cassandra Compaction is a process of reconciling various copies of data spread across distinct SSTables which is performed as a background activity.
While regular compactions are an integral part of any healthy Cassandra cluster, the way that they are configured can very significantly depend on the way a particular table is being used. It is important to take some time when designing a Cassandra table schema to think about how each table will be used, and therefore the Cassandra Compaction strategy would be most effective.
Although it is possible to change compaction strategies after the table has been created but it will re-write all of the data in that table using the new Compaction Strategy.
- Stores recent writes in memory (Memtable).
- When enough writes have been made, Cassandra flushes the Memtable to disk. Data on disk is stored in relatively simple data structures called Stored String Tables (SSTable). At the most simplified level, an SSTable could be described as a stored array of strings.
- Before writing a new SSTable, Cassandra merges and pre-sorts the data in the Memtable according to Primary Key.
- The SSTable is written to disk as a single contiguous write operation. SSTables are immutable. If data is updated regularly, Cassandra may need to read from multiple SSTables to retrieve a single row.
- Compaction operations occur periodically to re-write and combine SSTables. Compactions prune deleted data and merge disparate row data into new SSTables in order to reclaim disk space and keep read operations optimized.
- SizeTiered: When multiple SSTables of a similar size are present.
- Leveled
- DateTiered
- TimeWindow
The Maximum represents the highest number of SSTables (across all nodes over roughly the last five minutes) that were involved in a read request.