CloudHealth shows the EC2 instance start action as successful but the instance still stays in the stopped state.

Products

• CloudHealth

Issue/Introduction

We can see an error below for the instance when reviewing the change audit records.

"Amazon EC2 Instance i-0xxxxnxxxxxx1 has been modified: reason changed from 'User initiated' to 'Server.InternalError' "

Resolution

CloudHealth change audit records to track EC2 instance state changes, logging initiators (e.g., user or policy) and failure reasons like the shift from "User initiated" to "Server.InternalError" for instance i-0xxxxnxxxxxx1.

This message—"Amazon EC2 Instance i-0xxxxnxxxxxx1 has been modified: reason changed from 'User initiated' to 'Server.InternalError'"—means AWS accepted the start request but encountered an internal backend issue, overriding the user attribution and reverting the instance to stopped.​​

Review records at https://apps.cloudhealthtech.com/audit_records/

The "Server.InternalError" indicates an issue within AWS infrastructure, not a user-caused problem.

This error can occur during

> instance state transitions (like start, stop, or reboot) and is often transient. 
> Temporary issues with AWS services or infrastructure.​
> Disk or volume problems, like a corrupted or incompatible EBS volume.​
> Problems with instance launch, especially if the root or attached volumes are not correctly configured.​​

As recommended by  AWS, if the error persists, review the instance's system logs, checking attached EBS volumes, and try to relaunch the instance. Often, simply retrying the operation or launching a new instance resolves the issue.​

https://docs.aws.amazon.com/autoscaling/ec2/userguide/ts-as-instancelaunchfailure.html