EKS restore failure due to invalid instance type

Summary

Creating a new EKS cluster during restore may fail if an invalid instance type is selected when defining the restore job. Follow the instructions in this section to select the correct instance type for your new EKS cluster.

Step-by-Step

1. When restoring from a backup of discovered EKS cluster (i.e. a cluster with a linked AWS account), CloudCasa will automatically select the instance type for you based on the instance type used by the original cluster. However, for clusters which are registered manually, the instance type must be provided. Selecting an invalid instance type will result in a job failure with messages like the following in the activity log:

   2022-11-25 10:44:49 UTC: Creating node group  for AWS EKS cluster Achyut-Rest-25-11-2022
   2022-11-25 10:44:49 UTC: [a1.medium] is not a valid instance type for requested amiType AL2_x86_64
   2022-11-25 10:44:49 UTC: Internal operation failed
   

2. When selecting an appropriate instance type for your node groups, it is important to take the AMI type into account as not all AMI types will be supported for all instance types. If restoring from an existing EKS cluster, the AMI type for your managed node group can be obtained from the AWS console by navigating to the Elastic Kubernetes Service service page and then selecting your EKS cluster. Here you will see a list of nodes shown for your EKS cluster along with the instance type for each node.

3. If you are not restoring from an existing EKS cluster, or if you just want to change your instance type during the restore, refer to the AWS documentation Choosing an Amazon EC2 instance type to ensure that you are selecting a valid instance type when defining your restore job.