Hi, everyone.

We're running several Slony clusters with many sets inside each cluster. They were all set up using the exact same 
scripts. Out of all those sets only one is continuously experiencing problems.

Every couple of days some changes don't make it to the slave node. This includes inserts and deletes (possibly updates 
also, but I can't say for sure). There are currently around 1000 changes missing on the slave node. Recreating the set 
or rebuilding the cluster fixes things, but only for a couple of days.

The replication does not hang! New changes for that set are being replicated without any problems. It's just changes 
from a certain time in the past that are missing. Unfortunately I can't determine if they were all made during the same 
transaction or not.

We're using Slony 2.0.2 with PG 8.3.7 on Centos 5.3 64bit. Database encoding is UTF8 and cluster locale is set to C. 
There is nothing in sl_log_1/2 tables. Grepping PG and Slony logs (on both servers) for "ERROR" does not reveal anything.

Does anyone have any suggestions what else I could try to find out what is causing this?

Regards,
Aleksander