Examples of using DLPAR and CoD resources

Memory allocation process is similar.

It is important to remember that once PowerHA® SystemMirror® acquires additional resources for an application controllers, when the server moves again to another node, it takes the resources with it, that is, the LPAR node releases all the additional resources it acquired, and remains with just the minimum.

The configuration is an 8 CPU frame, with a two-node (each an LPAR) cluster. There are 2 CPUs available in the CoD pool, that is through the CoD activations. The nodes have the following characteristics:

The following application controllers are defined in separate resource groups:

Example 1: No CPUs are allocated at application controller start, some CPUs are released at server stop

Current® configuration settings:

• Node1 has 3 CPUs allocated.
• Node2 has 1 CPU allocated.
• The free pool has 4 CPUs.

PowerHA SystemMirror starts application controllers as follows:

• Node1 starts AS2, no CPUs are allocated to meet the requirement of 3 CPUs. (3 CPUs is equal to the sum on Node1's LPAR minimum of 1 plus AS2 desired amount of 2).
• Node1 stops AS2. 2 CPUs are released, leaving 1 CPU, the minimum requirement. (Since no other application controllers are running, the only requirement is Node1 LPAR minimum of 1).

Example 2: Failure to allocate CPUs due to resource group processing order

Current configuration settings:

• Node1 has 3 CPUs allocated.
• Node2 has 1 CPU allocated.
• The free pool has 4 CPUs.

PowerHA SystemMirror starts application controllers as follows:

• Node1 starts AS1, no CPUs are allocated since the requirement of 2 is met.

  Node1 starts AS3, 3 CPUs are allocated to meet the requirement of 6. There is now 1 CPU in the free pool.

• Node1 attempts to start AS2. After Node 1 has acquired AS1 and AS3, the total amount of CPUs Node1 must now own to satisfy these requirements is 6, which is the sum of Node1 LPAR minimum of 1 plus AS1 desired amount of 1 plus AS3 desired amount of 4.

  Since AS2 minimum amount is 2, in order to acquire AS2, Node1 needs to allocate 2 more CPUs, but there is only 1 CPU remaining in the free pool and it does not meet the minimum requirement of 2 CPUs for AS2. The resource group with AS2 goes into error state since there is only 1 CPU in the free pool and CoD use is not allowed.

Example 3: Successful CoD resources allocation and release

Current configuration settings:

• Node1 has 3 CPUs allocated.
• Node2 has 1 CPU allocated.
• The free pool has 4 CPUs.

PowerHA SystemMirror starts application controllers as follows:

• Node1 starts AS3, 2 CPUs are allocated to meet the requirement of 5.
• Node1 starts AS2, 2 CPUs are allocated to meet the requirement of 7. There are now no CPUs in the free pool.
• Node1 starts AS1, 1 CPU is taken from CoD and allocated to meet the requirement of 8.
• Node1 stops AS3, 4 CPUs are released and 1 of those CPUs is put back into the CoD pool.

Example 4: Resource group failure (the minimum for the server Is not met, but the LPAR maximum for the node Is reached)

Current configuration settings:

• Node1 has 1 CPU allocated.
• Node2 has 1 CPU allocated.
• The free pool has 6 CPUs.

PowerHA SystemMirror starts application controllers as follows:

• Node2 starts AS3, 4 CPUs are allocated to meet the requirement of 5. There are now 2 CPUs in the free pool.
• Node2 attempts to start AS2, but AS2 goes into error state since the LPAR maximum for Node2 is 5 and Node2 cannot acquire more CPUs.

Example 5: Resource group fallover

Current configuration settings:

• Node1 has 3 CPUs allocated.
• Node2 has 1 CPU allocated.
• The Free pool has 4 CPUs.

PowerHA SystemMirror starts application controllers as follows:

• Node1 starts AS2, no CPUs are allocated to meet the requirement of 3.
• The resource group with AS2 falls over from Node1 to Node2.
• Node1 stops AS2. 2 CPUs are released, leaving 1 CPU on the LPAR, the minimum requirement for the node.
• Node2 start AS2, 2 CPUs are allocated to meet the requirement of 3.