When a Database fail-over occurs in a DAG, the database will remain active on the new active node, until this node fails, or "ActiveServer", on the database, is changed manually.
This is a problem for me, on site servers. If they are rebooted, it is important, that the database becomes active on the site server, when it's up again.
Therefore, I have made a script, that I run from "Scheduled tasks"
Remember, that the user, that executes the script, must be member of "organization administrators"(or corresponding exchange group), local admin group for the server and have "log on as a batch job rights"
Make a scheduled task, that runs "C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe c:\PSscripts\MoveDB.ps1"
I have set the script to run every morning, but could also be a startup script, but the there should be put a delay in the script.
The Powershell script, will be run from the powershell.exe, that does not know the Exchange PS commands, so I start the script with adding the exchange snapin.
The 2 liner script:
Add-PSSnapin microsoft.Exchange.Management.PowerShell.E2010
Get-MailboxDatabase DBname | where { $_.server -notlike "siteserver-servername" } | Move-ActiveMailboxDatabase -ActivateOnServer 'siteserver-servername' -MountDialOverride 'None' -Confirm:$false -TerminateOnWarning | fl >> "C:\PSscripts\MoveDB2.LOG"
Subscribe to:
Post Comments (Atom)
Posts
Seems like I'm not the only one with this issue.
ReplyDeleteMS is implementing a script, that can balance the distribution of active database copies across a database availability group (DAG), in SP1 for Exchange 2010
Nice
Look under: "Active Mailbox Database Redistribution"
http://blogs.technet.com/scottschnoll/archive/2010/04/10/new-high-availability-features-in-exchange-2010-sp1.aspx
Thanks for this. Super annoying that there isn't an automatic method to return all databases back to their activate preference server. SP1 can't come soon enough!
ReplyDelete