Pool Considerations

In appropriate environments, you can add or remove group members from a storage pool or you can combine storage pools.

A single-array storage pool provides fault isolation. Volumes whose pools are on one array keep data placement simple and snapshots truly local. All host operating systems identified in the Validated Configuration Matrix and array data protocols support single-array pools.

A multi-array pool lets you consolidate capacity and scale performance. Volumes striped across multiple array pools can have larger capacity, pool-wide resources, and more even growth with less rebalancing later. Some host operating systems support multi-array pools.

Consider the following points when planning whether and how to implement storage pools.
  • Operating system on host machines

    Windows, ESX, and Linux hosts for which a supported Connection Manager is available can use multi-array storage pools. Any hosts for which a supported connection manager is unavailable require single-array pools.

  • Disk speed
  • Disk capacity
  • Network bandwidth
  • Application using the storage pool
  • Load balancing requirements
  • Volumes that will reside on the storage pool
If you are planning on pooling arrays across dissimilar models, consider the following caveats.
  • Coupling a lower performance model array with a higher performance model could degrade the performance of the pooled system.
  • Performance can be affected when you merge arrays into a single pool when the arrays have different performance caps, cache configurations, and storage capacity.
  • Consider using a multi-array pool to migrate data from an older array to a newer array, as this method can be less disruptive than using replication to migrate data.

Storage pools can include some or all of the arrays in a group. When you add an array to a group, you must also add this array to one of the existing storage pools for that group or you can create a new storage pool and add the array to the new pool.

The first implementation example shows a group that includes three HPE arrays that are all in the same group in the default pool.


  1. Array 1
  2. Array 2
  3. Array 3
  4. Default pool
  5. Multi-array group that includes arrays 1, 2, and 3
The second implementation example shows two pools that allow a specific application, such as Microsoft Exchange, to use volumes that are spread across two of three arrays in the group, but not use the storage on a third array. In this scenario, the storage on the array labeled 3 is in the default pool. To configure this example, remove the arrays labeled 1 and 5 from the default pool, and then create an Exchange storage pool by adding only the arrays labeled 1 and 5 to the pool.


  1. Array 1
  2. Exchange pool
  3. Array 3
  4. Default pool
  5. Array 5
  6. Multi-array group that includes arrays 1, 3, and 5
The third implementation example shows adding a new array that is labeled 6 to the group that is assigned to the default storage pool. In this scenario, you have other alternatives. One alternative is to move array 6 to the previously created Exchange pool. Another alternative is to create a new storage pool for array 6 and leave only array 4 in the default pool.


  1. Array 1
  2. Exchange pool
  3. Array 3
  4. Default pool
  5. Array 5
  6. Array 6
  7. Multi-array group that includes arrays 1, 3, 5, and 6
The fourth implementation example shows having one group with four HPE arrays, where each array is in a separate storage pool.


  1. Array 1
  2. Exchange pool
  3. Array 3
  4. Default pool
  5. Array 5
  6. SQL Server pool
  7. Array 7
  8. VDI pool
  9. Multi-array group that includes arrays 1, 3, 5, and 7