 Archive style Transaction log drive filled during point-in-time database restoration. ~Wei Opagen on April 19, 2011 | Version 2 |
Recent customer experience relating to problems following the point-in-time restoration of a database that caused the transaction log drive to be filled by the backup software, resulting in the Domino server crashing (an expected result).
|
So, you're using Archive type transaction logging in conjunction with a Certified Backup Solution when during a routing point-in-time restore the transaction log fills up and the Domino Server crashes,
what next?, what is the best approach to recover from this?, this article shares some of the lessons learned and some useful related information. The Domino environment consists of a pair of clustered Domino v8.5 servers, used predominantly for mail but with some applications. DAOS is enabled on both servers.
Transaction logging is enabled on both servers; Server A uses Circular tx log with 4Gb of allocated disk space and Server B uses Archive tx log with 90Gb of allocated disk space.
The backup product is Symantec Net Backup v6.5.6 Both servers have a full backup preformed weekly on a Saturday
Server B (archive tx log) has incremental backups of the transaction log nightly; this is essential so that transaction log extents that are marked as archive-able are backed up and then flagged as ready for re-use by the transaction logger. On average there was approx 10Gb of transaction log data being backed up per night,
|