Exchange drive loses space weekly

Ok, I think I'm on to something. It seems that my mdbdata folder may be where the space is being eaten. Any way to limit this or clean up this folder?

Try and run a compact and repair on the Exchange databases. The only problem is that if your exchange database is 2Gb, then you need more than 2Gb of space on the drive that it resides on to do the job. Reason is that it copies the database and rebuilds it, much like Access 2000 does when you run a compact and repair on an Access database. To cut down the size of your mailbox, you can take steps to clear out old, unused or hidden mailboxes. Also, if your users don't rely havily on recovering deleted items, then you can goto the properties of the Private Information Store and set the Deleted Item Retention Time to 0 days instead of the default which I think is 7. After a few days, the database should have reduced itslef to a controllable level, which we hope will enable you to run te compact and repair.

Since this happens on a cycle, I think your problem is transaction logs. How often do you run Exchange backups and what type of backups are they (normal, incremental, differential)?

Transaction logs are created as a normal part of Exchange operations. Each log file is almost exactly 5MB in size (5120KB to be exact). As one log file is filled with transactions, a new one is created. Transaction logs accumulate until either a normal (full) or incremental backup is accomplished. Depending on the number of users you support and the amount of email that is sent and received, these logs can accumulate at an alarming rate. I would be willing to bet that your mdbdata folder is getting filled with .log files. You may need to change your backup strategy (more often or use incremental instead of differential). BTW, if you attempt to just delete ANY of these .log files, chances are you will corrupt your Exchange data stores.

Hope this helps.

Tim Walsh is right, we had the same problem, we had over 2000 files with .log, we cured this by unclicking the log option. We have had no problem since.

This question was closed by the author