SCF Reference Manual for the Storage Subsystem (G06.28+, H06.05+, J06.03+)
7 Managing Disks
This chapter describes managing disks. Disks share the object type of DISK with virtual disks. See
“Configuring and Managing Virtual Disks” (page 143). For information about disk load balancing,
see “Disk Load Balancing” (page 118). This chapter describes:
• “Managing Disks” (page 96)
• “Starting a Disk” (page 97)
◦ “Reviving a Mirrored Disk” (page 98)
• “Stopping a Disk” (page 101)
• “Resetting a Disk” (page 103)
“Resetting One Disk” (page 103)◦
◦ “Resetting a Group of Disks” (page 104)
• “Sparing a Defective Sector” (page 104)
• “Correcting Doubly Allocated File Extents” (page 106)
• “Replacing the Bootstrap Program” (page 107)
• “Enabling and Disabling File Opens on a Disk” (page 109)
• “Swapping Processors for a Disk” (page 109)
• “Changing the Active Path for a Disk” (page 112)
• “Managing Encrypted Disk Drives” (page 116)
• Handling Power Failures
Managing Disks
When managing disks, review these special considerations:
• “Managing the System Disk” (page 96)
• “Recovering From the Loss of a Path to a Disk” (page 96)
• “Troubleshooting Disks” (page 97)
Managing the System Disk
You cannot stop access to the system disk because important processes have their object files and
swap files on it. Therefore, you cannot stop the last path to the system disk. You must take special
precautions when performing some procedures:
• If you have to replace the disk bootstrap program on the system disk, make sure it is mirrored.
See “Changing Two Nonmirrored Disks Into a Mirrored Volume” (page 79).
• Do not issue a “STOPOPENS Command” (page 293) on the system disk.
Recovering From the Loss of a Path to a Disk
If the X fabric or Y fabric fails, internal disk paths using the failed fabric go down. After repairing
a fabric failure, these failed disk paths are not automatically restarted. They remain HARDDOWN
until you restart the disk process. The storage subsystem never attempts to use the failed path,
which creates a potential single point of failure.
To restore the paths to all disks that use the repaired fabric, see “Resetting a Group of Disks”
(page 104).
96 Managing Disks