Oracle Cloud Infrastructure Documentation

Skip to main content

Autoscale Configuration Fields

The following tables describe each autoscale configuration field in detail. For information on creating an autoscale configuration, see Creating an Autoscale Configuration.

Fields for both Vertical and Horizontal Autoscaling


Field	Description
Autoscale configuration name	Enter a name for this configuration.
Cluster admin password	Enter the admin password for the cluster. (You assign an admin password when you create a cluster.)
Node type	Select the type of node. The available options are as follows: Worker: Select this option to add block storage and compute to your cluster. Compute only worker: Select this option to add only compute to your cluster.
Autoscale type	The available options are as follows: Horizontal: Select this option to add or remove nodes from the cluster. This option is available only if you select Compute only worker node type. Vertical: Select this option to change the shape of the cluster nodes.
Trigger type	Begins the autoscale based on the metrics of the nodes provided when configuring the autoscale configuration schedule.
Performance Metrics	Begins the autoscale based on the metrics of the nodes provided when configuring the autoscale configuration schedule. This is dependant on the metric trigger type. Performance metrics include: CPU utilization of the clusters nodes Memory utilization of the cluster nodes

Fields for Vertical Autoscaling


Field	Description
Scale-up rule	This field appears when you select the vertical autoscale type and metrics for the trigger type. The scale-up rule sets the conditions for scaling up the cluster (use a larger compute shape). If you selected CPU utilization in the Performance metrics field, enter the following details: Average CPU threshold percentage: Set a percentage of average CPU utilization, so that when the CPU operates at that average percentage (or higher) for the minimum duration in minutes, the cluster is scaled up. Minimum duration in minutes: Set a duration, so that when the average CPU threshold percentage operates for that duration, the cluster is scaled up. If you selected Memory utilization in the Performance metrics field, enter the following details: Average MEMORY threshold percentage: Set a percentage of average Memory utilization, so that when the memory operates at that average percentage (or lower) for the minimum duration in minutes, the cluster is scaled down. Minimum duration in minutes: Set a duration, so that when the average memory threshold percentage operates for that duration, the cluster is scaled down. If you selected a Flex shape, then also specify the following details: Maximum limits: Enter the maximum number of OCPUs and memory the cluster can have. Step size: Specify the number of OCPUs and memory to be added to the cluster each time the autoscale is triggered. For example, if the scale-up rule is: Threshold percentage = 80% Minimum duration = 30 minutes and the shapes of the worker nodes are VM.Standard2.4, when the CPU utilization averages 80% or more for 30 minutes, the worker nodes are scaled up to VM.Standard2.8. For Flex shapes, the shape of worker nodes changes depending on the OCPU and memory specified.
Scale-down rule	This field appears when you select the vertical autoscale type and metrics for the trigger type. The scale-down rule sets the conditions for scaling down the cluster (use a smaller compute shape). If you selected CPU utilization in the Performance metrics field, enter the following details: Average CPU threshold percentage: Set a percentage of average CPU utilization, so that when the CPU operates at that average percentage (or lower) for the minimum duration in minutes, the cluster is scaled down. Minimum duration in minutes: Set a duration, so that when the average CPU threshold percentage operates for that duration, the cluster is scaled down. If you selected Memory utilization in the Performance metrics field, enter the following details: Average MEMORY threshold percentage: Set a percentage of average Memory utilization, so that when the memory operates at that average percentage (or lower) for the minimum duration in minutes, the cluster is scaled down. Minimum duration in minutes: Set a duration, so that when the average memory threshold percentage operates for that duration, the cluster is scaled down. If you selected a Flex shape, then also specify the following details: Minimum limits: Enter the minimum number of OCPUs and memory the cluster can have. Step size: Specify the number of OCPUs and memory to be removed from the cluster each time the autoscale is triggered. For example, if the scale-down rule is: Threshold percentage = 20% Minimum duration = 30 minutes and the shapes of the worker nodes are VM.Standard2.8, when the CPU utilization averages 20% or less for 30 minutes, the worker nodes are scaled down to VM.Standard2.4. For Flex shapes, the shape of worker nodes changes depending on the OCPU and memory specified. If you selected Memory utilization in the Performance metrics field, enter the following details:
Scheduled-based vertical condition	This field appears when you select scheduled-based vertical autoscale type and schedule for the trigger type. For schedule-based vertical autoscaling, enter the following details: Days: Enter the days of the week that you want the condition to run. Time: Enter a trigger time for the condition to begin. For example, 9:00 a.m. Time Optional: Enter second trigger time on the same day. For example, 4:00 p.m. Target shape: Select a target shape from the list. If you select a flex shape, you must customize the following resources: Number of OCPUs Amount of memory (GB) Target shape Optional: Select a second trigger target shape. For example, if the schedule-based Vertical condition is: Days = Monday, Tuesday, Friday Time = 9:00 a.m. Target shape = VM.Standard.2.8 Time Optional = 4:00 p.m. Target shape Optional = VM.Standard.2.4 In the example, the first trigger is set for 9:00 a.m. with a target shape of VM.Standard.2.8. The second trigger is set for 4:00 p.m. with a target shape of VM.Standard.2.4.

Fields for Horizontal Autoscaling


Field	Description
Scale-out rule	This field appears when you select the horizontal autoscale type and metrics for the trigger type. The scale-out rule sets the conditions for scaling out the cluster (add cluster nodes). Note: Horizontal autoscaling applies only to compute only worker nodes. If you selected CPU utilization in the Performance metrics field, enter the following details: Average CPU threshold percentage: Set a percentage of average CPU utilization, so that when the CPU operates at that average percentage (or larger) for the minimum duration in minutes, the cluster is scaled out. Minimum duration in minutes: Set a duration, so that when the average CPU threshold percentage operates for that duration, the cluster is scaled out. Maximum number of nodes: Enter the max number of nodes the cluster can have. Step size: Specify the number of nodes to add during a scale-out event. If you selected Memory utilization in the Performance metrics field, enter the following details: Average memory threshold percentage: Set a percentage of average Memory utilization, so that when the Memory operates at that average percentage (or larger) for the minimum duration in minutes, the cluster is scaled out. Minimum duration in minutes: Set a duration, so that when the average Memory threshold percentage operates for that duration, the cluster is scaled out. Maximum number of nodes: Enter the max number of nodes the cluster can have. Step size: Specify the number of nodes to add during a scale-out event.
Scale-in rule	This field appears when you select the horizontal autoscale type and metrics for the trigger type. Note: Horizontal autoscaling applies only to compute only worker nodes. The scale-in rule sets the conditions for scaling in the cluster (use a smaller compute shape). Average CPU threshold percentage: Set a percentage of average CPU utilization, so that when the CPU operates at that average percentage (or larger) for the minimum duration in minutes, the cluster is scaled out. Minimum duration in minutes: Set a duration, so that when the average CPU threshold percentage operates for that duration, the cluster is scaled out. Maximum number of nodes: Enter the max number of nodes the cluster can have. Step size: Specify the number of nodes to add during a scale-out event. If you selected Memory utilization in the Performance metrics field, enter the following details: Average memory threshold percentage: Set a percentage of average Memory utilization, so that when the Memory operates at that average percentage (or larger) for the minimum duration in minutes, the cluster is scaled out. Minimum duration in minutes: Set a duration, so that when the average Memory threshold percentage operates for that duration, the cluster is scaled out. Maximum number of nodes: Enter the max number of nodes the cluster can have. Step size: Specify the number of nodes to add during a scale-out event.
Conditions	This field appears when you select scheduled-based horizontal autoscale type and schedule for the trigger type. Note: Scheduled-based horizontal autoscaling applies only to compute only worker nodes. For schedule-based horizontal autoscaling, enter the following details: Days: Enter the days of the week that you want the condition to run. Time: Enter a trigger time for the condition to begin. For example, 9:00 a.m. Time Optional: Enter second trigger time on the same day. For example, 4:00 p.m. Number of nodes: Enter the target number of nodes at first trigger. Number of nodes Optional: Enter the target number of nodes at second trigger. For example, if the schedule-based Horizontal condition is: Days = Monday, Tuesday, Friday Time = 9:00 a.m. Number of nodes = 10 Time Optional = 4:00 p.m. Number of nodes Optional = 1 In this example, 10 nodes are available between 9:00 a.m. and 4:00 p.m. on the specified days. However, when the time reaches 4:00 p.m. the nodes drop to one and remain at one until another trigger is reached.