The following bug has been logged on the website:
Bug reference: 19029
Logged by: Thadeus Anand
Email address: thadeus@rmkv.com
PostgreSQL version: 17.6
Operating system: Ubuntu 24.04.2
Description:
We have seven servers distributed geographically. We publish about 62 tables
through logical replication, with six subscribers. We were hugely affected
by the memory allocation bug after upgrading to 17.5. We had to suspend the
replication and resort to manually updating the tables.
So we updated to 17.6 almost immediately. Now the logical replication works
perfectly, but we notice the replication slot folders increasing in size
quite dramatically, even when there is no lag found in the subscribers.
We also have a replica set up for streaming replication, and that works fine
and that replication slot folder size is normal. So the issue is only with
logical replication.
Our wal_keep_size is set to 0, and my max_slot_wal_keep_size is set to 20GB.
I'm convinced this must be a bug, as it started happening only after the
minor version upgrade which was done three days ago. But if you have any
other suggestion, please help.
Restarting the subscriber does not help. Dropping the subscription clears
the replication slot, but that isn't a solution.
Now we are dropping and recreating the subscriptions every night, hoping we
will get a patch/solution soon.
If you need any further data, please ask and I will provide.
