checkpoint-starvation · shipping · also SQLite WAL checkpoint starvation

Checkpoint starvation

Checkpoint starvation is sustained inability to finish or reset a database log checkpoint because the required progress conditions keep being obstructed.

Related Observatory article: Checkpoint starvation on hesela.com

In SQLite WAL, retained reader snapshots can constrain backfill and recycling while later writes continue. [1] Our two-case observation separates a pinned transaction from an idle open connection; it is not a production failure-rate study. [4] A PASSIVE result can report busy=0 with fewer checkpointed frames than current log frames. [2]

Relations

related

database-checkpoint, write-ahead-log, sqlite-wal-end-mark

Limitations

• A large WAL file alone does not identify starvation; file length can remain high after progress. [1]

Citations

1. SQLite: Write-Ahead Logging, sections 2, 6 and 11 (accessed 2026-10-14)
2. SQLite: PRAGMA wal_checkpoint, three-column result (accessed 2026-10-14)
3. SQLite WAL checkpoint starvation: success is not completion (accessed 2026-10-14)
4. Hesela: measured SQLite WAL-reader observations and limits (accessed 2026-10-14)

JSON: /data/terms/checkpoint-starvation.json