LatencySensitivity

LatencySensitivity
LatencySensitivity

Specification of the latency-sensitivity information.

The latency-sensitivity is used to request from the kernel a constraint on the scheduling delay of the virtual CPUs or other resources. This allows latency-sensitive applications(e.g. VOIP, audio/video streaming, etc.) to run in a virtual machine which is configured to use specific scheduling latencies and to be scheduled with low latency.

The kernel does not provide any guarantee that it will meet the latency-sensitivity requirement of a virtual machine CPU or other resources but it will always accept the latency-sensitivity value provided.

AllOf
This class requires all of the following:
JSON Example
{
    "_typeName": "string",
    "level": "string",
    "sensitivity": 0
}
level
Required

Enumeration of the nominal latency-sensitive values which can be used to specify the latency-sensitivity level of the application.

In terms of latency-sensitivity the values relate: high>medium>normal>low.

Possible values:

  • low: The relative latency-sensitivity low value.

  • normal: The relative latency-sensitivity normal value.

    This is the default latency-sensitivity value.

  • medium: The relative latency-sensitivity medium value.

  • high: The relative latency-sensitivity high value.

  • custom:

    Deprecated as of vSphere API Ver 6.0. Value will be ignored and treated as "normal" latency sensitivity.

    The custom absolute latency-sensitivity specified in sensitivity property is used to define the latency-sensitivity.

    When this value is set to level the sensitivity property should be set also.

Possible values are : low, normal, medium, high, custom,
integer As int32 As int32
sensitivity
Optional

Deprecated as of vSphere version 5.5, this field is deprecated.

The custom absolute latency-sensitivity value of the application.

This value will be used only when the latency-sensitivity level property is is set to custom. It is ignored in all other cases.

The unit of this value is micro-seconds and the application is more latency sensitive when this value is smaller. For example, if the absolute latency-sensitivity is 2000us, the kernel will try to schedule the virtual machine in a way so that its scheduling latency is not more than 2ms.