Compaction, SSTables, Memtables, Tombstones, and Storage Engine Basics
Understand how Cassandra stores data under the hood and why storage engine behavior directly affects performance and maintenance.
Inside this chapter
- Why Internal Storage Matters
- Write Path Components
- Tombstones and Their Cost
- Compaction Strategy Awareness
Series navigation
Study the chapters in order for the clearest path from beginner Cassandra concepts to advanced distributed operations. Use the navigation at the bottom of each page to move through the full series.
Why Internal Storage Matters
Students can start with CQL, but long-term Cassandra success depends on understanding storage internals. SSTables, memtables, commit logs, compaction, and tombstones all shape performance and operational behavior.
Write Path Components
- Commit log for durability
- Memtable for in-memory writes
- SSTables for immutable on-disk storage
- Compaction for merging and cleanup over time
Tombstones and Their Cost
Deletes and TTL expirations create tombstones. Too many tombstones can make reads slower and repairs heavier. This is a major practical Cassandra topic and one reason data lifecycle design matters so much.
Compaction Strategy Awareness
Different compaction strategies suit different workloads. Time-series workloads, high-update workloads, and mixed read/write systems may need different choices. Advanced teams tune these based on observed behavior instead of default assumptions alone.