Autoscaling
Autoscaling automatically adjusts the number of running instances based on demand to maintain performance and control costs.
Instructions
Last updated
Was this helpful?
Autoscaling automatically adjusts the number of running instances based on demand to maintain performance and control costs.
To scale the instance count based on the current load of the service there must be two scaling rule conditions added. One rule to increase the instance count on heavy load and one rule to decrease the instance count after the load is reduced again.

Select Custom autoscale
Type in an appropriate name in Autoscale setting name
Select Scale based on a metric to enable autoscaling based on a defined rule set
Set the instance limits in line with your business requirements:
Minimum
Maximum
Default
Select Add a rule once you're ready to configure your Autoscaling rules
We recommend the following settings of the increase and decrease rules. But please keep in mind that this can depend on your workload and needs to be monitored and optimized!
Enable metric divided by instance count
Choose Greater than as the Operator and set the threshold to 70 (percent)
Set the duration to 10 (minutes)
Ensure that the Time grain statistic is set to Average
Ensure that the Operation is set to Increase count by
Set the Cool down (minutes) to 15 (minutes)
Check the Instance count (1) that will be added to the current instances
Click Add when done

Select Add a rule to add an additional rule
Enable metric divided by instance count
Choose Less than as the Operator and set the threshold to 35 (percent)
Set the duration to 20 (minutes)
Ensure that the Time grain statistic is set to Average
Ensure that the Operation is set to Decrease count by
Set the Cool down (minutes) to 30 (minutes)
Check the Instance count (1) that will be removed from the current instances
Click Add when done

Last updated
Was this helpful?
Was this helpful?