locked
Adding drives to my RAID10 Logs partition RRS feed

  • Question

  • Hi,

    I have a Windows 2012 Failover Cluster with 2 nodes and 2 SQL 2012 Always-on Availability Groups.  Both nodes are a physical machine and each node is the primary for an AG.  On one of the nodes I need to add drive space to the Logs partition which is in RAID10.  Since it is in RAID10 I cannot simply add the drives to the Logs partition (not supported by the PERC H810 RAID Controller) but rather I have to backup the logs onto another drive then recreate the Logs partition with the new drives added. 

    My plan is to failover the AG that is primary from the node that the work needs to be performed on so that the other node then is the primary for both AGs.  At this point I'm not sure about the correct steps that need to be taken to move the log files onto a different drive (That the other node wont be aware of) and recreate the partition with the new drives added.  Any thoughts?

    I have already posted this question on the Windows Failover Cluster forum and they directed me here(http://social.technet.microsoft.com/Forums/en-US/a0325739-febe-4fb6-b254-6bda0d4d8549/adding-drives-to-my-raid10-logs-partition).

    Wednesday, June 19, 2013 1:43 PM

Answers

  • Hi John,

    Your setup is a bit confusing without a diagram. I'm assuming that you have a 2-node Windows Server 2012 Failover Cluster where each node is running a standalone SQL Server 2012 instance. The standalone instances are running Availability Groups and one is a primary to the other and vice versa. Am I understanding your configuration correctly? 

    There are several ways to add drive space - adding disks, creating a new partition or using mountpoints. Once you've added the disk space, you can remove the database that you want to move to the new drive from the Availability Group. You can drop the database and restore from backup, moving the files to the new drive. You can then add it back to the Availability Group. You can't just detach the database because it is in RECOVERY mode. 

    If you have several databases in the Availability Group and you think that dropping them, restoring and adding them back in the Availability Group will be time consuming, you can opt to use a disk imaging application that would take a snapshot of the entire disk before you increase the space. Make sure that you shutdown the SQL Server service prior to doing this. After you've added disk space, restore the disk image to a now larger disk partition. 

    These are just to options that you can take to approach this. There are several others like just creating a new disk partition or using mountpoints but they would all require removing the database from the Availability Group first before moving the files to a different location and, then, adding them back in.


    Edwin Sarmiento SQL Server MVP
    Blog | Twitter | LinkedIn

    • Marked as answer by Fanny Liu Thursday, June 27, 2013 9:43 AM
    Friday, June 21, 2013 10:29 PM