IBM General Parallel File System for AIX: Administration and Programming Reference

Replacing disks in a GPFS file system

Replacing an existing disk in a GPFS file system with a new one is the same as performing a delete disk operation followed by an add disk. However, this operation eliminates the need to restripe the file system following the separate delete disk and add disk operations as data is automatically moved to the new disk.

Under no circumstances should you replace a stopped disk. You need to start a stopped disk before replacing it. If a disk cannot be started, you will have to delete it using the -p option on the mmdeldisk command. See the IBM General Parallel File System for AIX: Problem Determination Guide and search for disk media failures for further information on handling this.

Replacing disks in an SP environment

To replace virtual shared disks in a file system, you must decide if you will:

1. Create new virtual shared disks via the mmcrvsd command.
2. Create new virtual shared disks with more than one disk via the procedures outlined in the IBM Parallel System Support Programs for AIX: Managing Shared Disks manual.
3. Select virtual shared disks no longer in use by another GPFS file system.

The virtual shared disk may then replace an old disk in the file system via the mmrpldisk command.

For example, in file system fs0 to replace the virtual shared disk gpfs19vsd with the existing virtual shared disk gpfs21vsd, which is no longer in use by another file system, allowing the usage to default to dataAndMetadata, assigning the new virtual shared disk to a different failure group, 3, and allow the nodes participating in the migration of the data to default to all, enter:

mmrpldisk fs0 gpfs19vsd gpfs21vsd::::3 -v no

The system displays information similar to:

Replacing gpfs19vsd ...
 
GPFS: 6027-531 The following disks of fs0 will be formatted on node k145n03:
    gpfs19vsd: size 4390912 KB
Extending Allocation Map
GPFS: 6027-1503 Completed adding disks to file system fs0.
GPFS: 6027-589 Scanning file system metadata, phase 1 ... 
  77 % complete on Wed Jul 12 17:33:58 2000
 100 % complete on Wed Jul 12 17:33:59 2000
GPFS: 6027-552 Scan completed successfully.
GPFS: 6027-589 Scanning file system metadata, phase 2 ... 
GPFS: 6027-552 Scan completed successfully.
GPFS: 6027-589 Scanning file system metadata, phase 3 ... 
GPFS: 6027-552 Scan completed successfully.
GPFS: 6027-565 Scanning user file metadata ...
   1 % complete on Wed Jul 12 17:34:12 2000
 100 % complete on Wed Jul 12 17:34:15 2000
GPFS: 6027-552 Scan completed successfully.
Done
mmrpldisk: 6027-1371 Propagating the changes to all affected nodes.
This is an asynchronous process.

See the mmrpldisk Command for complete usage information.

Replacing disks in an HACMP environment

In an HACMP environment, to replace a logical volume, you must decide if you will:

1. Create a new logical volume via the mmcrlv command.
2. Select a previously created logical volume no longer in use by another GPFS file system.

The disk may then replace an old disk in the file system via the mmrpldisk command.

Note:

When using Fibre Channel disks, single-node quorum is not supported.

For example, in the file system fs1 to replace the concurrent logical volume gpfslv04 with the newly created concurrent logical volume gpfslv08, indicate that it should be used for metadataOnly, belong to failure group 2, and have only the nodes which have the file system mounted participate in the migration of the data, enter:

mmrpldisk fs1 gpfslv04 gpfslv08:::metadataOnly:2 -N mount

The system displays information similar to:

Replacing gpfslv04 ...
 
GPFS: 6027-531 The following disks of fs1 will be formatted on node k145n03:
    gpfslv08: size 4390912 KB
Extending Allocation Map
GPFS: 6027-1503 Completed adding disks to file system fs3.
GPFS: 6027-589 Scanning file system metadata, phase 1 ... 
  77 % complete on Wed Jul 12 17:33:58 2000
 100 % complete on Wed Jul 12 17:33:59 2000
GPFS: 6027-552 Scan completed successfully.
GPFS: 6027-589 Scanning file system metadata, phase 2 ... 
GPFS: 6027-552 Scan completed successfully.
GPFS: 6027-589 Scanning file system metadata, phase 3 ... 
GPFS: 6027-552 Scan completed successfully.
GPFS: 6027-565 Scanning user file metadata ...
   1 % complete on Wed Jul 12 17:34:12 2000
 100 % complete on Wed Jul 12 17:34:15 2000
GPFS: 6027-552 Scan completed successfully.
Done
mmrpldisk: 6027-1371 Propagating the changes to all affected nodes.
This is an asynchronous process.

See the mmrpldisk Command for complete usage information.