Cassandra Compaction Debt Degrading Read Latency
warninglatencyUpdated Jan 10, 2026
When Cassandra's pending compaction tasks (cassandra.compaction.tasks.pending) continually climb without dropping, the cluster is writing faster than it can compact SSTables. This causes read amplification as more SSTables must be scanned per query, directly degrading read latency despite healthy-looking CPU and memory.
Sources
Technologies:
How to detect:
Monitor cassandra.compaction.tasks.pending for sustained growth without drops. Correlate with rising read latency P99 (cassandra.client.request.read.latency.99p) while P50 remains stable. Watch for increasing SSTable count.
Recommended action:
Increase compaction throughput settings, add disk I/O capacity, or reduce write rate. Consider tuning compaction strategy or adding nodes to distribute load.