Why should you care?

PostgresVerse Mumbai primary streams to Pune standby so regional outage fails over with seconds of lag not hours of backup restore.

See it live — copy this example

Run in pgAdmin or psql.

-- Primary: confirm sender status
SELECT application_name, client_addr, state, sent_lsn, write_lsn, flush_lsn, replay_lsn
FROM pg_stat_replication;

-- Standby: replay position
SELECT pg_is_in_recovery(), pg_last_wal_replay_lsn();

What happened?

• pg_stat_replication shows each standby catch-up stage — sent, written, flushed, replayed LSN.
• Standby pg_is_in_recovery true means read-only replica mode.

Practice next

1. Set wal_level=replica and max_wal_senders on primary.
2. Configure recovery on standby with primary_conninfo (lab guide).
3. Query both sides LSN gap.
4. Promote standby with pg_ctl promote and re-point apps.
5. Compare pg_stat_replication sync_state synchronous vs async.

Remember

WAL stream is continuous physical sync. Standby hot — readable if hot_standby on. Sync vs async trades durability for latency.

PostgresVerse geo replica

Reporting dashboards read Pune standby; writes stay on Mumbai primary.

Outcome: Primary CPU drops 25% during month-end reports.