I have a running instance on us-east which I can connect to. It is an EBS volume based on one of the ubuntu 10.10 server AMIs.

It was created by starting a default image, and then customizing it and then saving that as an AMI. I then terminated the original and ran an image of the AMI, and was able to connect.

At this point I used cloudy scripts "copy ami to different region" script (at https://cloudyscripts.com/tool/show/5) to copy it to us-west. This script copies the EBS

Once there, when I run it, I am unable to connect. I am trying to connect to it exactly the same way as I did in us-east. That is,


but this results in a timeout with error "bad file number"

I went back to the original image on us-east and restored it, and was still able to connect to it.

The instance is launched in security group: default which is configured this way in both regions:

protocol: tcp from port: 22 to port:22 source CDR

I really can't see any difference between the two setups - both are launched from an AMI, which should be the same one, but one works and the other doesn't. Any ideas?

  • Can you try running your SSH command with -v, and include that output?
    – Andrew M.
    Nov 16 '10 at 13:13
  • the only extra information is "attempt to connect timed out without establishing a connection" Nov 16 '10 at 13:47

Amazon now provides a way to copy an AMI from one region to another without manually copying EBS volumes.

This can be done using the AWS Management Console or by any of the SDKs. For the CLI, use aws ec2 copy-image.

Use this feature instead of the old custom script.


aha, it turned out that the instance was crashing upon restart due to kernel mismatch. I found out from checking the console output.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.