I have a nightly process that basically copies tables from a database source to a SQL Server database target. The process identifies the deltas between the source and target and stores those in a separate database. Those delta rows are then applied to the target database. The deltas are about 200 MBs a night.
I have a weekly full database backup scheduled with daily transaction log backups. Indexes are reorganized daily and rebuilt weekly. The backups are taken after the nightly processes and index jobs.
The full backup is about 25 GB compressed. I would expect the transaction log backup to be no more than 200 MBs compressed as well. However, the transaction log backups are generally in GBs and quickly become larger than the full backup.
The target database is in the Full recovery model. I also noticed that the transaction log grew to 255 GB even though the database was only 68 GB. I assume that the large transaction log backups are associated with the transaction log file growth. However, when the database was in simple recovery, the transaction log never exceeded its initial size of 10 GB.
Any idea what may be causing the transaction log to grow so quickly? Is the index maintenance job the cause of the issue? If so, what would the recommendation be to minimize the backup space needed?
