PostgreSQL I/O shows input/output metrics for a monitored PostgreSQL server. Use it to diagnose slow storage, checkpoint pressure, high read/write workload, and resource contention.
Workflow
- Open I/O.
- Review read, write, and timing indicators where available.
- Compare spikes with Activity and Queries.
- Use Waits to confirm I/O-related wait pressure.
How To Interpret PostgreSQL I/O
PostgreSQL I/O evidence helps explain whether storage activity is part of a performance problem. Heavy reads can come from missing indexes, large scans, reporting workload, cache churn, or queries that return more data than the application needs. Heavy writes can come from normal transaction volume, batch jobs, checkpoints, vacuum activity, or bulk loads.
Start with the time window. Match I/O spikes to user complaints, scheduled jobs, backups, imports, maintenance, or deployment activity. Then use PostgreSQL Queries to identify statements that may be driving reads or writes and PostgreSQL Waits to confirm whether sessions are actually waiting on I/O.
If host OS login is configured, operating system disk metrics can add context that PostgreSQL-only views may not show. See Host OS Login for the host metrics available through SSH or Windows host monitoring.
PostgreSQL I/O FAQ
Does high I/O mean the disks are too slow?
Not always. High I/O may be caused by query shape, table scans, cache pressure, maintenance tasks, or normal workload peaks.
What should I compare with I/O spikes?
Compare activity, queries, waits, checkpoints, backups, imports, and application release times.
Why are some storage metrics unavailable?
Availability depends on PostgreSQL version, extensions, permissions, host OS login, and whether the database is self-managed or cloud-hosted.