Compaction in cassandra is merging data from multiple SS tables , does it perform splitting data from one ss table into multiple ones ?
Hi @ashok.dcosta_187920: Compaction is the process of freeing up space by merging large accumulated data‐ files. So it is meant to merge smaller SST tables into a bigger one and not the other way around.
A compaction operation in Cassandra is performed in order to merge SSTables. SSTables are immutable, which helps Cassandra achieve such high write speeds. However, periodic compaction of these SSTables is important in order to support fast read performance and clean out stale data values. During compaction, the data in SSTables are merged: the keys are merged, columns are combined, obsolete values are discarded, and a new index is created.
For more details see How compaction works in Cassandra. Thanks!!
Thanks for being part of the community but I wanted to point out that these statements aren't always true and can be misleading to new users:
So it is meant to merge smaller SST tables into a bigger one and not the other way around.
SSTables are immutable, which helps Cassandra achieve such high write speeds.
In the process of this compaction does total number of SSTables decrease or increase? Having 600 SSTables for examples can be normal?
7 People are following this question.
