Change EC2 instance type from Intel to AMD
AWS Resource Type:
The AMD-based EC2 instances that AWS offers are typically about 10% cheaper than the equivalent Intel EC2 instances but provide a very similar level of performance. This FF identifies Intel EC2 instances that can be safely retyped to their AMD counterparts.
Specifically, the instance families that this FF will retype are:
T3 -> T3a, M5 -> M5a, C5 -> C5a, R5 -> R5a, M6i -> M6a, C6i -> C6a, and R6i -> R6a.
The FF requires a short downtime for each instance that is retyped because the instance must be restarted for the change to take effect.
Criteria for identifying the opportunity:
This FF considers the following EC2 instances:
- Only Intel instance family types deemed safe to change to their AMD counterpart: T3, C5, M5, R5, C6i, M6i, and R6i.
- Instances for which we do not have at least 30 days of CPU and IO throughput data are excluded.
- Instances with measured peak EBS throughput in the last 30 days that exceed the EBS bandwidth of the AMD counterpart are excluded.
- For instance, types T3, C5, M5, and R5 make sure CPU utilization in the last 30 days was less than 80% (due to a lower clock frequency of the AMD counterpart).
- Instance types that have reservations are excluded.
- Instance types that are not long-lived are excluded (e.g., part of Auto Scaling Group or EMR).
Potential savings (range in % on an annual basis):
Across the CloudFix user base, we have found that roughly 25% of EC2 spending is on-demand instances using Intel types that can be migrated to AMD. We can reduce around 10% of its costs.
What happens when the Fixer is executed?
The fixer does the following:
- Take an EBS snapshot while the instance is running.
- Stop the EC2 instance.
- Update the EBS snapshot now that the instance is stopped.
- Change the instance type to its AMD counterpart.
- Start the EC2 instance.
- If it started successfully, the fixer is done after setting up CloudWatch monitoring of EBS bandwidth and IOPS and network throughput for that instance. This will allow an automatic rollback if those metrics indicate regression in performance.
- if the instance failed to start, initiate an automatic rollback which essentially changes the type back to Intel and attempts to start the instance.
- If it still fails to start, initiate a manual rollback. There is no runbook for this manual rollback, but rather, the customer needs to repair and restart the instance (if necessary, the EBS snapshot can be used to create a new EC2 instance). This scenario is highly unlikely.
Is it possible to rollback once CloudFix implements the fixer?
Yes. The rollback changes the instance type back to the Intel type of the instance. If the automatic rollback fails, a manual rollback is initiated, during which the EBS snapshot can be used to build a new EC2 instance if necessary.
Can CloudFix implement the fix automatically once I accept the recommendation?
Does this fix require downtime?
Yes. The EC2 instance is rebooted.