- Storage Engines · System Design
LLearnViaNotes @learnvianotes· 2 min read Deep dive into LSM Trees, B+ Trees, SSTables, Bloom Filters, Compaction algorithms, and MemTables driving modern databases.
LSM Trees (Log-Structured Merge-Trees)
Storage Engine Comparison Matrix
| Feature | B+ Tree (e.g., PostgreSQL, InnoDB) | LSM Tree (e.g., Cassandra, RocksDB) |
|---|
| Write Bottleneck | Random I/O (Page splits, dirty buffer updates) | Sequential I/O (WAL append + MemTable) |
| Read Bottleneck | Minimal (Direct index lookup O(log N)) | Compaction background overhead + Multi-file lookups |
| Write Amplification | High (Full 16KB page rewrite for small updates) | Moderate to High (Compaction rewriting SSTables) |
| Space Efficiency | Fragmentation over time | Compact & Compressed block SSTables |
Free for students and teachers
Write notes like this one
Cards for every idea, quizzes and flashcards, read aloud — on the web and your phone.
Something wrong with this note? Open it to report