Databases
Storage engines, indexing, ACID and isolation levels.
9 notes · 61 min total
- 1Core DatabasesAudioThe four storage types almost every design uses: relational, NoSQL, blob storage and search, and when each one fits.10 min
- 2ACIDAudioWhat transactions are and how atomicity, consistency, isolation and durability keep data correct through crashes and concurrency.9 min
- 3Isolation LevelsAudioThe read anomalies concurrent transactions cause, the four SQL isolation levels that prevent them, and how serializable is implemented.9 min
- 4In-Memory vs Disk StorageAudioHow disk-based, in-memory and LSM-based databases each use RAM and disk, and what that means for speed and data safety.5 min
- 5Database IndexingAudioWhy indexes turn full table scans into fast lookups, and when to use primary, secondary, clustered, composite and covering indexes.7 min
- 6Hash IndexesAudioHow hash indexes give O(1) exact-match lookups, why they can't do ranges, how collisions are handled, and when to choose them over B-trees.4 min
- 7Read-Heavy Indexing: B-TreesAudioHow a B-tree finds any row in a few disk reads, how inserts and deletes keep it balanced, and why every extra index slows writes.4 min
- 8Write-Heavy Indexing: LSM TreesAudioHow LSM trees turn random writes into sequential ones using a memtable, SSTables, bloom filters, compaction, tombstones and a WAL.7 min
- 9Indexing Trade-offsAudioWhat every index costs in write speed and storage, when not to add one, and which engines use B-trees, LSM trees or other structures.6 min