PostgreSQL admin_shutdown
The server terminated the client connection because PostgreSQL was stopped, restarted, promoted, or explicitly instructed to end sessions.
FATAL: terminating connection due to administrator commandAlso searched as: admin_shutdown · terminating connection
What it means
This is a server-side lifecycle event rather than proof of a network fault. Clients should discard the broken session, reconnect to a healthy endpoint, and retry only operations known to be safe.
Typical trigger conditions
- Planned restart or deployment
- High-availability failover or promotion
- Operator termination of sessions
- Managed-database maintenance
First diagnostic checks
Correlate server events
Inspect PostgreSQL, service-manager and provider events at the exact timestamp.
Confirm endpoint role
Verify that the client is reconnecting to the current writer or appropriate read endpoint.
SELECT pg_is_in_recovery(), inet_server_addr(), inet_server_port();
Assess transaction outcome
Do not blindly replay non-idempotent writes when commit status is uncertain.
Version and platform notes
The SQLSTATE is stable; failover and maintenance evidence is platform-specific.
Follow PostgreSQL failures and fixes
One useful weekly email with new error references, tools, integration notes and production lessons. No daily noise.