Monitoring, Errors, and Unexpected Behavior FAQ

How do I monitor my systems?

Amazon EC2 provides basic monitoring. You can use DescribeInstances to check whether an instance appears to be running.

For more advanced monitoring, consider Amazon CloudwWatch. Amazon CloudWatch runs a monitoring services that collects raw measurement data, such as CPUUtilization (percentage of Amazon EC2 compute units used by an instance) and DiskWriteBytes (number of bytes written in a minute). For more information, refer to the Amazon CloudWatch Developer Guide.

Why can't I "talk" to my instances?

There are a few common reasons for broken connectivity to your instance.

Amazon EC2 changes the state of your instance to running after your operating system starts booting. Depending on your AMI, there will be a delay before the instance is fully set up and functional.

If your instance has been running for several minutes, you verify you authorized the appropriate access to your host through the Amazon EC2 firewall. If you have launched your instances without specifying a security group, the default group is used. Permissions on the default group are very strict and disallow all access from the Internet and other groups. You will need to modify the permissions of your default group or set up a new group with appropriate permissions. For more information, see Network Security Concepts

If this doesn't solve your issue, make sure you authorized port 22 and try to open an SSH connection with verbose output. Use the man page for the exact syntax of your system, but the command is likely to be similar to ssh -vv root@[hostname]. This output is very useful if you are posting to the forum.

Why did my instance terminate immediately after launch?

Launch errors can be the result of an internal error during launch or a corrupt Amazon EC2 image. Internal errors are rare, as we actively test for and isolate suspect hosts. Consult the DescribeInstances operation for details on why your instance failed to launch.

	Note
	The `ec2-describe-instances` command line tool does not provide this information. Use the `-v` flag to read the detailed SOAP response and get detailed information.

You can also attempt to launch the image again. If this proves to be a persistent problem (especially with a shared image), post to the AWS forums.

I ran shutdown from within an ssh session, but my instance still shows up as running when I query it with DescribeInstances and I can't shell into it.

To shut down an instance, use the TerminateInstances call (ec2-terminate) on the command line. You can also use shutdown -h, but must verify the instance shut down using the DescribeInstances call.

Why are my instances stuck in a pending state (or a shutting-down state)?

This situation is rare and might be the result of a software error or misconfiguration.

We actively monitor for this; please contact us if it occurs.

Why do I get an "AuthFailure: User is not AMI creator" error when I try to register an image?

Make sure that you are using the correct user ID and certificate to create and upload the image. You must use the same ID and certificate to register the image with Amazon EC2.