Repository navigation
torznab_indexer_latency grows unboundedly — CleanupOldLatency exists and is tested but never called #2376
Description
On my instance: 54,781 rows accumulated in 7 days ( Likely a consequence of this: the table's indexes are disproportionately large relative to its data even accounting for the row count — Suggested fix: call Steps to reproduce
EnvironmentRelevant logsChecklist
|
Replies: 2 comments
|
Confirmed on current main: On the index size question, the schema rules out the row count as the cause: CREATE TABLE torznab_indexer_latency (
id INTEGER GENERATED BY DEFAULT AS IDENTITY PRIMARY KEY,
indexer_id INTEGER NOT NULL REFERENCES torznab_indexers(id) ON DELETE CASCADE,
operation_type TEXT NOT NULL,
latency_ms INTEGER NOT NULL,
success INTEGER NOT NULL,
measured_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);Every indexed column is fixed width apart from a short For wiring the cleanup, there is already a pattern for exactly this in the codebase, |
Solved in #2457