Runaway Transaction Logs

I use the term runaway for want of a more accurate technical description.

I’ve seen transaction logging runaway a few times over the past few years, most recently just this weekend.

The initial symptoms that alert you to this issue vary, databases are offline, transaction log volumes are full, transaction logging is out of control or mailboxes are full of NDR’s, but by and large the issue is the same.

In my recent customers case transaction log disk space was being consumed at upwards of 1Mbyte per second, and a 30GByte log volume that should last a week is filling up in just over 8 hours, between backups, and databases are being dismounted.

The most recent case is on Exchange 2003, but I’ve also seen it with Exchange 2007 and Exchange 2010.

Continue reading “Runaway Transaction Logs”