How to fix System error in PostgreSQL

Review the PostgreSQL server logs to identify the specific system error. Logs are typically located in /var/log/postgresql/ on Linux systems.

grep -i "system error" /var/log/postgresql/postgresql.log
grep -i "error" /var/log/postgresql/postgresql.log | tail -20

The detailed error message will indicate whether the issue is related to disk I/O, permissions, file access, or other system resources.

Check that the PostgreSQL data directory has sufficient free space. PostgreSQL requires well over 20 MB free under /usr/local and at least 25 MB on the partition containing the data directory.

df -h /var/lib/postgresql
du -sh /var/lib/postgresql/main

If disk space is low, delete old backups, logs, or temporary files, or expand the partition.

# Free up space by rotating old logs
postgresql-logrotate

The PostgreSQL data directory must be owned by the postgres user with permissions 0700 (owner access only).

ls -ld /var/lib/postgresql/main

The output should show something like: drwx------ ... postgres postgres ...

If permissions are incorrect, fix them:

sudo chown -R postgres:postgres /var/lib/postgresql/main
sudo chmod 700 /var/lib/postgresql/main

If PostgreSQL crashed and failed to shut down cleanly, the postmaster.pid file may prevent restart. Check if the process is truly gone before removing the file:

cat /var/lib/postgresql/main/postmaster.pid
ps aux | grep <PID from file>

If the process is not running, safely remove the stale PID file:

sudo rm /var/lib/postgresql/main/postmaster.pid
sudo systemctl restart postgresql

PostgreSQL will attempt automatic recovery on the next start.

Use system monitoring tools to check for resource problems that could trigger system errors:

# Check disk I/O and errors
iostat -x 1

# Check memory usage
free -h
watch -n 1 free

# Check for filesystem errors
dmesg | grep -i error
fsck -n /dev/sdXX  # (read-only check, safe to run)

High I/O wait, memory pressure, or filesystem error messages in dmesg indicate hardware or system configuration problems.

If the logs show "database system was not properly shut down; automatic recovery in progress", PostgreSQL will handle recovery automatically on restart. Allow it to complete without interruption:

sudo systemctl start postgresql
sudo tail -f /var/log/postgresql/postgresql.log

Wait for logs to show the database is accepting connections before running queries. Recovery time depends on the number of incomplete transactions.