Temp write pressure
Query audit
Postgres work_mem check for spill-risk review
Database Optimization Tool reviews PostgreSQL temp write and query evidence before teams consider work_mem changes, keeping all output evidence-focused and advisory.
Sort and hash spill signals
Query-level evidence
Settings stay under team control
work_mem evidence starts with spills
The check reviews temp block writes, query runtime, call counts, row counts, and spill-heavy query patterns to identify where memory-related investigation may be useful.
Query context before tuning
A work_mem change can raise memory pressure under concurrency. Database Optimization Tool points teams toward EXPLAIN (ANALYZE, BUFFERS), query shape review, and controlled testing.
Evidence-gated recommendations
Findings are tied to collected PostgreSQL evidence such as pg_stat_statements and temp write pressure. Unsupported AI guesses are not treated as production instructions.
evidence-focused and advisory
The product does not change work_mem, rewrite queries, create indexes, or tune database settings. AI-assisted notes stay advisory only and require human review.
FAQ
Frequently asked questions
These answers describe the product focus: careful database evidence, clear findings, and team-approved next steps.
Can Database Optimization Tool change work_mem?
No. It is evidence-focused. It reports evidence and diagnostics so engineers can decide whether controlled testing is warranted.
Do temp writes always mean work_mem is too low?
No. Temp writes can come from query shape, joins, aggregation, sorting, estimates, or expected analytical work. The audit treats them as context-dependent signals.
Is a global work_mem increase recommended?
No. The page avoids broad memory-change advice because concurrency and workload shape can make a global increase risky.
Does the AI output replace EXPLAIN review?
No. AI notes are advisory only. Engineers should verify spill causes with PostgreSQL evidence such as EXPLAIN (ANALYZE, BUFFERS).