Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I have an ec2 server that was running for several months without any problems, but I recently had to reboot due to an unknown reason.

Now, it appears that it is having trouble starting up, possible because of a script I asked it to run on startup. Looking at the system log from the ec2 management console, I find a scary message, that I am suspicious may be the cause.

There are alternative theories, but it seems to be hanging in the middle of running my script.

Is there a way for me to boot up the server at a different run-level, or in some type of rescue mode to bypass this startup script and fix whatever is wrong?

[  149.318024]   

[  149.318054]   ***************************************************************

[  149.318060]   ***************************************************************

[  149.318066]   ** WARNING: Currently emulating unsupported memory accesses  **

[  149.318071]   **          in /lib/tls glibc libraries. The emulation is    **

[  149.318076]   **          slow. To ensure full performance you should      **

[  149.318081]   **          install a 'xen-friendly' (nosegneg) version of   **

[  149.318085]   **          the library, or disable tls support by executing **

[  149.318090]   **          the following as root:                           **

[  149.318095]   **          mv /lib/tls /lib/tls.disabled                    **

[  149.318100]   ** Offending process: nginx (pid=803)                        **

[  149.318105]   ***************************************************************

[  149.318109]   ***************************************************************

[  149.318117]
share|improve this question

2 Answers 2

I'm fairly sure you can't boot an EC2 instance to rescue mode or single user mode.

If it's an EBS backed instance then you can attach the EBS root volume to another instance and access it that way.

Failing that you may have to use AWS premium support.

share|improve this answer

Here's an article I wrote describing how to access the root disk of an EBS boot instance that doesn't boot:

Fixing Files on the Root EBS Volume of an EC2 Instance
http://alestic.com/2011/02/ec2-fix-ebs-root

You could disable the startup script and then try another boot.

The tls nosegneg warning is not likely to be causing any problems related to boot.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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