Total time
High total time usually points to repeated user impact. It can matter more than one rare, very slow statement.
Problem path
A PostgreSQL slow query audit starts by ranking pressure, not guessing fixes.
Slow query work should connect query statistics, index evidence, table health, and business workflow impact before anyone changes production SQL.
Mean time
High mean time shows heavy individual queries that may block checkout, reporting, booking, or admin workflows.
Temp writes
Temporary files can expose sort, join, or memory pressure that deserves EXPLAIN review before tuning.
Disk reads
Read pressure can point to missing index candidates, inefficient joins, stale statistics, or cold-cache behavior.
Index evidence
Index recommendations should connect to a hot query shape and be tested before production creation.
Table health
Bloat, dead tuples, and vacuum drift can make otherwise reasonable queries behave poorly under load.
No automatic production changes
The audit path should produce a prioritized review list. If the problem is not yet PostgreSQL-specific, start with broad SQL performance tuning or index optimization first. Query rewrites, index creation, and configuration changes still need owner approval and rollout planning.
Need a specialist PostgreSQL audit?
If slow query review needs hosted workflow, account handling, pricing, or transaction support, use the dedicated PostgreSQL specialist service instead of this category guide.