Exchange Errors due to Undetected Page Removal while B-tree Split Operation

Microsoft Exchange Server storage groups are consisted of mailbox store and public folder store. Each mailbox store is further made up of database set comprising of two files: Priv.edb and Priv.stm. Similarly, each public folder store database set contains Pub.edb and Pub.stm files. For any reason if Information Store databases get corrupt, you might experience symptoms like slow performance, abrupt behavior while normal operations and most often, JET errors. If an updated data backup is available, you are recommended to restore from it to solve these problems. In case no valid copy exists, try running built-in ESEUTIL /P utility or safe EDB Repair products.

For instance, when using Microsoft Exchange Server, you may observe any of the below symptoms with information store databases:

1. You run ESEUTIL /D to attempt offline defragmentation of the database, but fail with -1601(JET_errRecordNotFound) or -1603(JET_errNoCurrentRecord) error

2. Error -1069 error (JET_errVersionStoreOutOfMemory) while normal database operations

3. Application log reports an error with Event ID: 1101

4. You run ESEUTIL /G /V /X in Exchange Server 5.5 and receive JET Error -1206 (JET_errDatabaseCorrupted).

Cause

The Information Store service might violate access because of encountered corrupt data.

When trying to insert a record in B-tree, the Exchange engine might find a record with same primary key value, marked for deletion. In such cases, it undeletes the node by updating its content with that of new record and might even change the size of record. Now, if this size is too large for the page, the database engine performs b-tree split operation. The information store may become unresponsive or data might get corrupt if Exchange Server tries to re-latch the pages while the record has been actually removed by information store cleanup thread.

Solution

You can recover from such problems by applying these measures:

1. Execute database consistency checker to check for corrupted and/or orphaned LVs

2. For correcting corrupted and orpahned LV errors, you need to run ESEUTIL /P /X followed by offline defragmentation

3. Perform offline defragmentation if only orphaned LVs are reported

4. In case of corrupted LVs, you should run ESEUTIL /P /X followed by offline defragmentation

If applying these measures doesn't fix the database corruption problem, you should restore from the complete database backup or use a EDB Repair Tool in case of valid backup unavailability. EDB Repair utilities are graphically rich tools incorporated with high-end scanning algorithms to repair and restore damaged Exchange databases at a safe location.

Stellar Phoenix Mailbox Exchange Recovery is an advanced utility that employs high-end scanning algorithms to repair a damaged Exchange database. It supports Exchange Server 5.5, 2000 and 2003. It is a safe EDB Repair Tool with impressive features. The tool extracts user mailboxes in .pst file format.