I'm aware that usually the historical replication data should delete from this folder when it passes the set amount of time for being there.
Over the past few months however I've noticed this isn't happening for one of our publications, meaning we have to go in and delete the historical data manually to ensure plenty of space is available on the server.
We're running SQL 2008, using Snapshot Replication which runs daily. I have an inkling of what it might be down to but I'm hoping someone can confirm my thoughts.
On the server in question ([server1]), we have two databases ([db1] and [db2]), both of which are replicated daily to [server2] via a schedule. There is also a third server that we have subscribed to the publication for [db1], but the synchronization for that is run on demand and not via a schedule.
In the repldata folder, the...