[scaling] section - AWS ParallelCluster

[scaling] section

Specifies settings that define how the compute nodes scale.

The format is [scaling scaling-name]. scaling-name must start with a letter, contain no more than 30 characters, and only contain letters, numbers, hyphens (-), and underscores (_).

[scaling custom] scaledown_idletime = 10

scaledown_idletime

(Optional) Specifies the amount of time in minutes without a job, after which the compute node terminates.

This parameter isn't used if awsbatch is the scheduler.

The default value is 10.

scaledown_idletime = 10

Update policy: The compute fleet must be stopped for this setting to be changed for an update.