Fleetmanagement, EC2 Auto Scaling.
The idea of EC2 Auto Scaling groups is to ensure automated the required compute capacity. Auto Scaling independently detect if a failure occurs in one or more instances and an automated replacement takes place. The Maximum, Minimum and the Desired capacity will defined initially e.g. via the Management Console. If for example two instances fail, EC2 Auto Scaling Group is able to rebuild two new instances as a replacement. The replacement is also well balanced between the Availability Zones (Above AZ1 vs. AZ2).
The capacity could be set as a conditional and/or scheduled capacity demand in relation to the calendar and scaling in or out is possible. Typically a Auto Scaling group is placed beneath a Load Balancer or behind a queue. The Load Balancer delivers traffic into an Auto Scaling group and we could use measurements from Load Balancer to manage better the Auto Scaling group. Optionally the usage of CloudWatch supports the monitoring of the Auto Scaling metrics in a centralized component as well.
Return to main post of AWSOMESDAYS 2019 agenda?
Click here https://cutt.ly/GeYaUS9