Managing storage policies

A storage policy is a group of parameters that define how redundant a VM volume must be and where it needs to be located. A storage policy can also be used to limit the bandwidth or IOPS of the volume. These limits help customize the allocation of cluster resources between the virtual machines. It is also needed to provide predictable performance levels for virtual machine disks.

When you deploy the compute cluster, you select parameters for the default storage policy, to be automatically created together with the cluster. The default policy cannot be deleted or renamed. By default, it is applied to uploaded images and base volumes created from these images.

A base volume is created from a source image when you deploy a VM. It is not used directly by a VM, but all volumes that a VM actually uses (which are listed on the Volumes tab) are in fact deltas (differences) from the base volume. It is important to keep base volumes available as VM volumes depend on them. For that, you need the default storage policy to enforce multiple replicas.

If the storage cluster does not have enough nodes to enable multiple replicas (not recommended), you can adjust the default storage policy once you add more nodes to the storage cluster. It will be applied to the images and base volumes that were created with the default policy.

You can create and manage policies for the following storage locations:

  • Local storage cluster. A compute volume will reside on the Virtuozzo Storage cluster with the chosen redundancy settings, including a tier, failure domain, and redundancy mode.
  • External storage. A compute volume will reside on an attached third-party storage compatible with OpenStack. Virtuozzo Hybrid Infrastructure does not provide data redundancy for such storage systems.