53100: disk_full

What does it mean?

The server could not write to a file because the disk is full. No space is left on the disk holding the data directory, WAL or temporary files.

Why does it happen?

Data growth, uncontrolled WAL accumulation (archiving or a replication slot falling behind), large temporary files (huge sorts), bloated log files, or table/index bloat.

How to fix it

Inspect disk usage and clean up unnecessary files. Check WAL accumulation: pg_stat_replication and pg_replication_slots (drop unused slots; lagging ones retain WAL). Do not run VACUUM FULL or REINDEX while the disk is full: because they rewrite the table they need extra space equal to the table size and can make things worse; plain VACUUM also (almost) never returns space to the operating system. NEVER delete files in the pg_wal directory by hand. First free space by resolving the WAL/slot, log and temp sources; review the placement of logs and temp files. As a permanent fix, increase disk capacity and set up monitoring/alerting.