Hello,
I have a two-node Windows 2008 R2 cluster on a MD3200 SAS SAN. Several days ago the event log reported "The file system structure on the disk is corrupt and unusable. Please run the chkdsk utility on the volume SQLBACK.". SQLBACK is one of the cluster disks allocated to SQL Server and it is about 600GB in size. The disk is used for database transaction logs (LDF) and for backup files (TRN, DIF and BAK). The disk is currently marked as dirty.
This SQL cluster is a critical system for my company and the window I have to take it down is very small. From what I have read I need to run chkdsk but doing so could take hours for the size of that disk. I am also concerned if the cluster fails over for any reason it will also run chkdsk during the mounting of the drives because the dirty flag is set. I have read you can stop the chkdsk from running during mounting by changing a setting on the cluster disk. I have also moved all database's transaction logs (LDF) and I have changed the database backups (trn, dif and bak) to go to another cluster disk. At this point nothing from SQL Server is using that cluster disk resource.
In order to avoid SQL downtime, I want to temporarily remove the disk from SQL Server list of dependent drives. Then take the drive offline for maintenance and then do a chkdsk to repair whatever is ailing the drive. The chkdsk can run for how ever long it needs because SQL will still be up. Once the drive is fixed I will bring it back online and then add it back to SQL Server as a dependent drive. If all goes well I will move all the files back onto this drive I moved off earlier.
I have googled and googled and I am unable to find any posted steps to accomplish this task. Is what I want to accomplish viable and if so what are the steps to make this happen.
Thank You.