Re: How to measure replication lag in logical replication setup.

From: pgsql-admin-owner(at)postgresql(dot)org [mailto:pgsql-admin-owner(at)postgresql(dot)org] On Behalf Of rammohan ganapavarapu
Sent: Tuesday, November 07, 2017 9:52 PM
To: pgsql-admin(at)postgresql(dot)org
Subject: [ADMIN] How to measure replication lag in logical replication setup.

HI,

How to measure replication lag time in case of logical replication?

this query is returning empty

SELECT now() - pg_last_xact_replay_timestamp() AS time_lag;
time_lag
----------

(1 row)

this is my slot details

SELECT * FROM pg_replication_slots;
slot_name | plugin | slot_type | datoid | database | active | xmin | catalog_xmin | restart_lsn
-----------+--------+-----------+--------+----------+--------+------+--------------+-------------
slot_1 | | physical | | | t | 1983 | | 0/50805C8

In my case slot_type is physical, does it look ok or for logical replication the slot_type should be logical?

Thanks,
Ram

I compare confirmed_flush_lsn in PG_REPLICATION_SLOTS on the Publisher with pg_current_wal_lsn():

SELECT confirmed_flush_lsn, pg_current_wal_lsn(), (pg_current_wal_lsn() - confirmed_flush_lsn) AS lsn_distance
FROM pg_catalog.pg_replication_slots
WHERE slot_name = 'subscriber_1';

slot_type for logical replication is logical.
Slot was created “automatically”, when I created “subscriber_1” subscription.

Regards,
Igor Neyman