Replication is in "initializing" status for a long time (looks like lack of documentation)

[evgeniagusakova]

Hi!

I set up replication for a big database (~1.7T) and it is in initializing state for a long time (~1 week) Database is in-use, I can't stop INSERTS/UPDATES/DELETES

#  SELECT * FROM pglogical.show_subscription_status();
 subscription_name |    status    |  provider_node  |                             provider_dsn                   |                slot_name                 | replication_sets | forward_origins
-------------------+--------------+-----------------+------------------------------------------------------------+------------------------------------------+------------------+-----------------
 subscription      | initializing | my_db__provider | host=192.168.30.230 dbname=my_db_production user=pglogical | pgl_my_db9012fe1_my_db99b54_subscription | {default}        | {all}
(1 row)

# select * from pglogical.local_sync_status;
 sync_kind | sync_subid | sync_nspname | sync_relname | sync_status | sync_statuslsn
-----------+------------+--------------+--------------+-------------+----------------
 d         | 2875150205 |              |              | d           | 0/0

I do not see any network/disk/OS/other issues, and after checking documentation I did not find any way to check initializing progress.

• Are any ways to check initializing progress?
• Are any ways to get initializing ETA?

[kamilski81]

@evgeniagusakova Did you find any answer to this? I am getting the same issue.

[tahseenbaig]

@evgeniagusakova Did this get resolved? could you please elaborate a bit? I am in the same situation.

[mjuvette]

I am having the same issue. Please an idea how to resolved this. The status of the subscription remain as initializing and no errors in the logs

[tahseenbaig]

@mjuvette , I ended up creating multiple subscriptions. First I added all tables to one replication set ( rep_set_1) then I removed the big chunky ones along with their sequences. Created subscription on target for rep_set_1 Next, I created individual replication sets for each of the big tables ( rep_set_tabA), added their sequences to it as well, and got that replicated by creating corresponding subscriptions for them (sub_tabA) on target.

hope that helps!