Everyday I learn something new. A couple of days ago I posted about a problem I had with Always On and a full transaction log. Finding root cause came from an article from Pinal Dave stating basically everything we knew what happened.
To recap the problem, transaction log backups were not truncating the log and the replica’s were not synchronizing. So come back to what we know was going on. There was an index maintenance job running during the window in which we seen the large transaction log backups.
So with those facts stated, lets talk about what we know about transaction logs in general. We know that each and every transaction in the database is recorded in the log. With that said we can expect large transactions obviously take more log space, especially rebuilding indexes. This explains the growth in the log. So since the re-index had to replicate to the replica’s this caused our log backups to stop truncating and eventually they just filled the mount point to the point in which they could not grow.
Always be learning folks!