Overview of custom-sized provisioning
Custom-sized (or variable-sized) provisioning has more flexibility than fixed-sized provisioning and is the traditional storage-based volume management strategy.
To create custom-sized volumes on a storage system, an administrator creates volumes of the desired size from individual array groups. These volumes are then individually mapped to one or more host ports as logical units (LUs).
Custom-sized provisioning provides advantages in the following three scenarios:
In fixed-sized provisioning, when several important files are located on the same volume and one unimportant file is being accessed, users cannot access the important files because of logical device contention. If the custom-sized feature is used to divide the volume into several smaller volumes and I/O workload is balanced (each file is allocated to a different volume), then access contention is reduced and access performance is improved.
In fixed-sized provisioning, all of the volume's capacity might not be used. Unused capacity on the volume will remain inaccessible to other users. If the custom-sized feature is used, you can create smaller volumes that do not waste capacity.
Applications that require the capacity of many fixed-sized volumes can instead be given fewer large volumes to relieve device addressing constraints.
The following illustrates custom-sized provisioning in an open-systems environment using standard volumes of independent array groups:
Disadvantages
A disadvantage is that manual intervention can become costly and tedious. For example, to change the size of a volume already in use, you must first create a new volume larger (if possible) than the old volume, and then move the contents of the old volume to the new volume. The new volume is then remapped on the server to take the mount point of the old volume, which is then retired.
When to use custom-sized provisioning
Use custom-sized provisioning when you want to manually control and monitor your storage resources and usage scenarios.