posts - 104 , comments - 115 , trackbacks - 0

Reseeding the passive node of a DAG/CCR

Running a CCR or DAG is a great way to have a server resilient database. Relying on an asynchronous copy of the log files (Each log file gets copied over only when the store has released it) it copies over the log files from the active node to the passive node so it can be played in to the passive nodes database. But whilst being a great failover solution both solution are also quite high-maintenance, CCR more so than a DAG. Running in to a situation where the CCR or DAG is down because of to many missing log files is annoying at best, but manageable.

When that happens it will be required to reseed the passive node, a time consuming process but fairly easy.

Reseeding the passive Exchange 2007 node:

1. Logon to the passive node
2. Suspend the replication between two nodes by executing Suspend-StorageGroupCopy -identity “<servername>\Generic Storage Group”
3. Delete or move all .log, .chk, .edb files on the passive node (for that storage group: F:\Generic storage group)
4. Copy over the .edb from the active node to the passive node
5. Resume the replication between the nodes by executing Resume-StorageGroupCopy -identity “<servername>\Generic Storage Group”

This should reseed the log files and .chk files to the passive node. Once the reseeding is complete the storage group should show up as healthy in the console (or by using the Get-StorageGroupCopyStatus command.)

Reseeding the passive Exchange 2010 node:

1. Logon to the passive node
2. Suspend the replication between the two nodes by executing suspend-mailboxdatabasecopy –identity Mailboxdatabasename\servername
3. The copy status will change to failed and suspended.
4. Reseed the database by executing Update-MailboxDatabaseCopy -Identity "Mailbox Database \servername" –DeleteExistingFiles (add in –manualresume if you want to prevent the automatic replication resume)

How long either process will take depends on the network speed. It will, however, bring your CCR or DAG back in to a current state. If Exchange is still complaining about to many missing log files (it happens >_<) you can opt to enable circular logging, take a backup (important!) and disable circular logging. Use that as a last resort only!

Technorati Tags: ,,,,,,,,,,,
,,,,,,,,,,
,,,,,,

Print | posted on Tuesday, January 31, 2012 11:10 PM | Filed Under [ Exchange ]

Feedback

No comments posted yet.
Post A Comment
Title:
Name:
Email:
Comment:
Verification:
 

Powered by: