Storage configuration planning

Complete all storage configuration planning tasks before you configure your system environment.

Planning the clustered systems

Determine:

The number of hosts to use.
The number of I/Os per second between the hosts and nodes.

Planning the hosts

Host mapping provides a way for hosts to access specific logical units (LUs) within the storage systems. Determine:

For hosts that use an Internet Small Computer System Interface (iSCSI) over Ethernet connection, the iSCSI qualified name (IQN) of the host and the authentication credentials.
The names to assign to the hosts.
The volumes to assign to the hosts.

Planning the MDisks

Planning the storage pools

Necessary activities:

Determine the types of storage systems to use.
To create volumes with the sequential policy: plan to create a separate storage pool for these volumes, or ensure that you create these volumes before you create volumes with the striped policy.
Plan to create storage pools for the storage systems that provide the same level of performance, reliability, or both performance and reliability. For example, group all of the managed disks that are RAID 10 in one storage pool and all of the MDisks that are RAID 5 in another storage pool.
Plan the extent size of the storage pool. For example, a larger extent size increases the total amount of storage that the system can manage. A smaller extent size provides more fine-grained control of storage allocation. Extent size does not affect performance.
If you are attaching more than 4 PB of storage to a system, you are not able to use all this capacity with the default extent size of 1 GB. Run the following CLI command one time to enable support for larger extent sizes in the GUI.
```
chsystem -guiadvancedpool yes
```

Planning the volumes

An individual volume is a member of one storage pool and one I/O group. The storage pool defines which MDisks provide the back-end storage that makes up the volume. The I/O group defines which nodes provide I/O access to the volume.

Before you create a volume, determine:

Whether to create the volume in image mode from a managed disk that contains data that must be preserved.
Note: You can map the logical unit number (LUN) presented from an external storage system as an image mode managed disk. Then, you can migrate data from that volume onto the system.
The volume name.
The volume I/O group.
The volume storage pool. For example, different storage pools might have different performance characteristics, depending on the storage that is contained by that storage pool.
The volume capacity.
Whether to provide extra redundancy by mirroring the volume across storage pools. For example, you might use the volume mirroring feature to provide redundancy across storage pools.
Whether to create fully allocated volumes or use thin-provisioned volumes.
Whether the cache mode for the volume is either readwrite, readonly, or none. The default is readwrite.

Consider the effect that FlashCopy®, Global Mirror, Metro Mirror, and thin-provisioned volumes have on performance. The effect depends on the type of I/O, and is calculated by using a weighting factor.

Compressed volumes are a special type of volume where data is compressed as it is written to the volume, saving extra space. A separate license is not required with IBM Spectrum Virtualize to use the compression function.

You can use the analyzevdiskbysystem command to determine potential compression savings for your current environment. Single volumes can also be analyzed for compression savings by using the analyzevdisk command.