{"id":"no-compaction-rate-limiting","text":"Neither LSM implementation throttles compaction I/O throughput; a large merge will saturate disk bandwidth without regard to concurrent foreground reads or writes.","truth_value":"IN","source":"entries/2026/05/29/topic-bitcask-merge-window-scheduling.md","source_url":"","source_hash":"","justifications":[],"dependents":[],"metadata":{},"explanation":{"steps":[{"node":"no-compaction-rate-limiting","truth_value":"IN","reason":"premise"}]}}