Why should you care?

When PostgresVerse primary crashes mid-transfer, WAL replay ensures committed banking debits are not lost.

See it live — copy this example

Run in pgAdmin or psql.

SELECT pg_current_wal_lsn(), pg_walfile_name(pg_current_wal_lsn());
SHOW wal_level;
SHOW max_wal_size;

What happened?

• pg_current_wal_lsn() is the log position now.
• pg_walfile_name maps to on-disk segment.
• wal_level must be replica or logical for replication; max_wal_size controls checkpoint frequency.

Practice next

1. Run WAL position query before and after INSERT.
2. Note wal_level — logical needed for logical replication lessons.
3. Find pg_wal directory size on disk (do not delete manually).
4. CHECKPOINT; then compare LSN jump.
5. Query pg_stat_wal for wal_records and wal_write time.

Remember

WAL is durability and recovery backbone. LSN tracks byte position in log stream. Replication slots prevent premature WAL removal.

PostgresVerse failover drill

Standby promoted after primary AZ failure; WAL replay catches up last 2 seconds of commits.

Outcome: RPO measured in seconds, not lost business day.