Product Documentation

AutoScale

Apr 26, 2017
What are the prerequisites for setting up AutoScale?
For prerequisites for setting up AutoScale, see "Prerequisites".
Can the CloudPlatform AutoScale feature be used without a NetScaler appliance?
No. The NetScaler appliance is currently required for the AutoScale feature to work. If the CloudPlatform administrator configures AutoScale in a network that does not include a NetScaler appliance, CloudPlatform throws an error.
What happens if the AutoScale feature is used with a NetScaler release that does not support AutoScale?
If the AutoScale feature is used with a NetScaler release that does not support AutoScale, the CloudPlatform user interface throws an error. CloudPlatform also writes a message to the log file, indicating that the configured NetScaler does not support AutoScale.
What versions of CloudPlatform and NetScaler should I use to implement AutoScale?
For information about NetScaler releases that support AutoScale, see Supported Environment.
In a load balancing rule, can manually provisioned virtual machine instances coexist with instances provisioned by the AutoScale feature?
No. The CloudPlatform virtual machine group in a load balancing rule can contain only manually provisioned instances or only instances provisioned by the AutoScale feature. They cannot coexist.
Is there a limit on the number of virtual machine instances to which we can scale up by using AutoScale?
Yes. The CloudPlatform administrator specifies the maximum number of members to which the configuration can scale up. When the limit is reached, virtual machines are not provisioned even if the scale-up condition is satisfied. The upper limit prevents uncontrolled spawning of VMs due to misconfiguration of the AutoScale feature or unexpected load conditions.
Are AutoScale events observable?
The events generated for deploying or destroying virtual machines are observable. These events are logged in the NetScaler logs (ns.log) and in the CloudPlatform logs (management-server.log). However, you cannot observe the metric values collected by NetScaler monitors.
What metrics can be used in AutoScale policies?
In an AutoScale policy, you can use any metric that is exposed through SNMP, or any NetScaler statistics associated with the load balancing virtual server used in the AutoScale configuration. For example, you can use metrics associated with CPU, memory, or disk usage, and NetScaler metrics such as throughput or response time.
What should a CloudPlatform administrator do before performing maintenance tasks on a CloudPlatform network in which AutoScale is configured?
The CloudPlatform administrator should disable the AutoScale configuration from the CloudPlatform user interface. Disabling the AutoScale configuration temporarily disables any scale-up or scale-down events. However, disabling AutoScale for an application, in CloudPlatform, does not affect the ability of the NetScaler appliance to serve traffic to existing virtual machines.
With AutoScale configured, are any configured VM limits enforced on the user account?
The NetScaler appliance works in the context of an AutoScale user account. Therefore, any limits that the CloudPlatform administrator has imposed on the number of VMs that can be created by the account are automatically enforced when the NetScaler appliance attempts to create more VMs than are permitted.
Is AutoScale supported in a high availability (HA) NetScaler pair?
No. Currently, HA mode is not supported for AutoScale.