Environment:

Linux 2.6.32-358.14.1.el6.x86_64
Slony 2.2.2
postgresql 9.3.3

Issue:

At times we see an issue in the slony replication logs that says:
ERROR:  invalid input for txid_snapshot: "233477269:233478374:233477269,233477846,233477978,233478062,233478210,233478287,233478308,233478314,233478360,233478360,233478371,233478372"

As you can see there is a duplicate value of transaction id in in the txid_snapshot, here it is related to the txid: 233478360 that is appearing twice. The only way we have been able to get around it is to manually update the sl_event table on the master node. Then slony just goes on to replicate as normal. The question here is why would slony event log have a duplicate entry in the first case and why can it not get rid of the duplicates itself. 

Also, one thing to consider here is that this issue does NOT occur as a ripple effect from any slonik command that we might have used. It occurs randomly in the middle of the day and we get to know that slony is lagging behind with this error. Are there other people who have reported this issue as we have encountered this a couple of times now not just in version 2.2.2 but in previous version 2.1.x

Please let me know if there is any further information that you would require and I would be happy to provide the same.

Thanks