Commandless disk replacement

Commandless disk replacement automates the process of replacing a failed or bad drive with a new drive.

The commandless disk replacement feature helps in automating the process of replacing a failed drive with a new drive. The disk hospital begins moving data off that drive whenever the drive fails in preparation for replacement. When data is drained off the drive, or if the drive is undrainable or dead, the disk hospital marks the drive as replaceable. Then, the commandless disk replacement component runs the prepare for replacement operation on the drive. As a part of the prepare operation, the LED on the drive for replacement is turned on, indicating that it is ready for removal.

If the prepare operation fails, then use commands like the mmvdisk pdisk replace for mmvdisk-managed recovery groups or mmchcarrier for non-mmvdisk recovery groups to replace the bad drive manually. For more information, see Disk Replacement.

Identify the bad drive that needs to be replaced and verify that the replace LED is turned on. Next, remove the bad drive, and insert a new drive. The new drive is identified by commandless disk replacement within minutes, as defined by the configuration keyword nsdRAIDDiskDiscoveryInterval. Commandless disk replacement then runs the replace operation. This operation makes the new drive ready for use. If it runs successfully without any errors, then no further action is required.

If the replace operation on a new drive fails for any reason, error messages, indicating the failure, are logged in the mmfs.log file. You can then replace the drive manually by following the manual disk replacement procedure. For more information, see Disk Replacement.