Create a switchable independent disk pool

Before you attempt to do switchable independent disk pools, ensure that you have satisfied the hardware, software, communications, and physical planning requirements. See Plan for independent disk pools.

iSeries™ Navigator is the recommended interface for creating and managing independent disk pools. Wizards in the clusters and disk management components simplify the tasks and guide you through the process. For some disk management tasks, iSeries Navigator is the only option. Make sure you have fulfilled the iSeries Navigator requirements for disk management.

After an expansion unit (frame/unit) is configured as switchable, the disk units in these resources cannot be used in independent disk pools that span different Cluster Resource Groups (CRGs). Even when only one node is defined and no switching will actually be enabled, configuring that expansion unit as switchable is enough to cause this constraint to be enforced.

Using iSeries Navigator

To create a switchable independent disk pool using iSeries Navigator, do the following:
  1. Create a cluster. To use switchable independent disk pools, an iSeries cluster is required.
  2. Make your hardware switchable. If you have a stand-alone expansion unit or an IOP that contains disk units that are to be included in an independent disk pool, you must authorize the expansion unit or IOP to grant access to other nodes.
  3. Create a switchable hardware group. A switchable hardware group, also known as a device CRG, defines the switchable independent disk pool. This is what manages the switching of the device. This wizard takes you through the steps to create a new switchable hardware group. It will also guide you through the New Disk Pool wizard which will assist you in creating a new disk pool and adding disk units to it for the cluster.
    Note: If you had switchable software products which conform to specific iSeries Navigator cluster guidelines installed when you ran the New Cluster wizard in step 1, the New Cluster wizard may have already prompted you to create a switchable hardware group. If the New Cluster wizard did not detect that a switchable software product was installed, then you have not created the switchable hardware group.
  4. Print your disk configuration. Print your disk configuration to have in case a recovery situation occurs. Also, record the relationship between the independent disk pool name and number.
You have now created a switchable independent disk pool. To prepare it for use, do the following:
  1. Start switchable hardware group. Start the switchable hardware group to enable device resiliency for the switchable hardware group.
  2. Make a disk pool available. To access the disk units in an independent disk pool, you must make the disk pool available (vary on) the disk pool.
  3. Perform a test switchover. Before you add data to the disk pool, perform a test switchover on the switchable hardware group you created to ensure the configuration functions as you planned.

Using CL commands and APIs

To create a switchable independent disk pool using CL commands and APIs, do the following:

You can use CL commands and APIs for creating a switchable independent disk pool, however there are some tasks that require that you use iSeries Navigator.
  1. Create the cluster. Create the cluster with required node using the CRTCLU (Create Cluster) Command.
  2. Start the nodes that comprise the cluster. Start the nodes in the cluster using the STRCLUNOD (Start Cluster Node) Command
  3. Create the device domain. You must create the device domain for all nodes involved in switching an independent disk pool or set of independent disk pools using the ADDDEVDMNE (Add Device Domain Entry) command.
  4. Create the device descriptions. Device descriptions must be created on each node that will be in the cluster resource group (CRG). Use the CRTDEVASP (Create Device Description (ASP)) Command. On the command line in the character-based interface, enter CRTDEVASP. In the Resource Name and the Device Description fields, enter the name of the independent disk pool you plan to create.
  5. Create the cluster resource group. Create the device CRG with the nodes, their roles in the recovery domain, and independent disk pool device descriptions using the CRTCRG (Create Cluster Resource Group) Command.
  6. Make your hardware switchable. If you have a stand-alone expansion unit or an IOP that contains disk units that are to be included in an independent disk pool, you must authorize the expansion unit or IOP to grant access to other nodes (iSeries Navigator required).
  7. . Create the disk pool on the node that owns the disk units using the New Disk Pool wizard when the server is fully restarted. Make sure clustering is active before you start. Name the independent disk pool to match the device description resource name that you gave in step 3. As you add disk units, it is best to localize disk units in the same expansion unit or IOP. Do not spread the disk pool across more device parity sets than necessary.
  8. Print your disk configuration. Print your disk configuration to have in case of a recovery situation. See How to display your disk configuration in Backup and Recovery. Link to PDF Also, record the relationship between the independent disk pool name and number.

    You have now created a switchable independent disk pool. The remaining steps are required to prepare it for use.  

  9. Start the cluster resource group. Start the cluster resource group to enable device resiliency using the STRCRG (Start Cluster Resource Group) Command.
  10. Make the disk pool available. To access the disk units in an independent disk pool you must vary on the disk pool using the VRYCFG (Vary Configuration) command.
  11. Perform a test switchover. Before you add data to the disk pool, perform a test switchover to ensure the configuration functions as you planned. Use the CHGCRGPRI (Change CRG Primary) command.

You are now ready to populate the independent disk pool with directories and libraries. Before you do, be sure to read Independent disk pools with distinct databases.