How to fix Could not start WAL streaming in PostgreSQL

Connect to the primary PostgreSQL server and check if the replication slot exists:

SELECT * FROM pg_replication_slots;

Look for your replication slot in the output. If it's missing, the slot needs to be recreated. Note the exact slot name (case-sensitive).

Verify that streaming replication is enabled on the primary:

SHOW wal_level;
SHOW max_wal_senders;
SHOW max_replication_slots;

Ensure:
- wal_level is set to replica or logical (not minimal)
- max_wal_senders is greater than 0 (at least 2 for one standby plus backups)
- max_replication_slots is at least the number of standby servers you have

If changes are needed, edit postgresql.conf and restart the primary server.

If the slot does not exist from Step 1, recreate it on the primary server:

-- For physical replication (streaming replication)
SELECT pg_create_physical_replication_slot('your_slot_name');

-- For logical replication (if applicable)
SELECT pg_create_logical_replication_slot('your_slot_name', 'test_decoding');

Replace your_slot_name with your actual slot name. Use the exact same name in the standby's configuration.

On the standby server, check the recovery configuration file (recovery.conf in PostgreSQL 11 and earlier, or signal files in PostgreSQL 12+).

For PostgreSQL 12+, ensure these files exist:
- standby.signal (to enable standby mode)
- postgresql.conf with primary_conninfo pointing to the primary

Example primary_conninfo:

primary_conninfo = 'host=192.168.1.100 port=5432 user=replication password=your_password'

Ensure the replication slot name matches exactly (case-sensitive) with what exists on the primary.

After configuration is verified, restart the PostgreSQL standby server:

sudo systemctl restart postgresql
# or
sudo pg_ctl restart -D /path/to/data/directory

Monitor the logs for connection attempts and errors:

tail -f /var/log/postgresql/postgresql.log

Once the standby connects, verify replication is working on the primary:

SELECT usename, application_name, state, sync_state FROM pg_stat_replication;

You should see your standby in the results with a state of streaming. Monitor replication lag:

SELECT slot_name, slot_type, restart_lsn FROM pg_replication_slots;