VMware vSphere Storage I/O Control allows cluster-wide storage I/O prioritization, which results in better workload consolidation and helps reduce extra costs associated with over provisioning.

vSphere Storage I/O Control extends the constructs of shares and limits to storage I/O resources. You can control the amount of storage I/O that is allocated to virtual machines during periods of I/O congestion, so that more important virtual machines get preference over less important virtual machines for I/O resource allocation.

When vSphere Storage I/O Control is enabled on a datastore, the ESXi host monitors the device latency when communicating with that datastore. When device latency exceeds a threshold, the datastore is considered to be congested and each virtual machine that accesses that datastore is allocated I/O resources in proportion to their shares. Shares are set on a per-virtual machine basis and can be adjusted.

vSphere Storage I/O Control has several requirements, limitations, and constraints.

  • Datastores that are enabled with vSphere Storage I/O Control must be managed by a single vCenter Server system.
  • Storage I/O Control is supported on Fibre Channel-connected, iSCSI-connected, and NFS-connected storage. RDM is not supported.
  • Storage I/O Control does not support datastores with multiple extents.
  • Before using vSphere Storage I/O Control on datastores that are backed by arrays with automated storage tiering capabilities, verify that the storage array has been certified as compatible with vSphere Storage I/O Control. See VMware Compatibility Guide.
Table 1. Design Decisions on Storage I/O Control

Decision ID

Design Decision

Design Justification

Design Implication

SDDC-VI-Storage-004

Enable Storage I/O Control with the default values on all non-vSAN datastores.

Storage I/O Control ensures that all virtual machines on a datastore receive an equal amount of I/O.

Virtual machines that use more I/O access the datastore with priority. Other virtual machines can access the datastore only when an I/O contention occurs on the datastore.