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.

This is driving me crazy since hours now. I have done the setup of 5 instances in AWS (some micro, some large) and after a while when i stopped them and rebooted them many times one of them won't boot.

I digged the problem to the point where when dracut start and try to switch the root filesystem there is some errors with the disk mapping and the /dev/xvd* devices.

Logs from a working one :

udev: starting version 147 dracut: Starting plymouth daemon xlblk_init: register_blkdev major: 202 blkfront: xvde1: barriers disabled Changing capacity of (202, 65) to 14680064 sectors xvde1: detected capacity change from 0 to 7516192768 EXT4-fs (xvde1): mounted filesystem with ordered data mode. Opts: dracut: Mounted root filesystem /dev/xvde1 dracut: Loading SELinux policy

Logs from a non working one :

udev: starting version 147 dracut: Starting plymouth daemon xlblk_init: register_blkdev major: 202 blkfront: xvde1: barriers disabled EXT4-fs (xvde1): mounted filesystem with ordered data mode. Opts: dracut: Mounted root filesystem /dev/xvde1 dracut: Loading SELinux policy

I have tried to delete /etc/blkid/* but without luck.

If someone could help me on this i would be very thankfull

share|improve this question
    
Using RHEL6 as OS –  WebOfMars Aug 23 '13 at 13:54
    
How do you know that it doesn't boot? As opposed to sshd not starting up. In case you have some other services (eg., httpd) - try to access those. Also, detach EBS volume from non working instance and attach (as /dev/xvdf or similar) to a working distance. Take a look at /etc/ssh/sshd_config for some "funny" lines at the end of the file. Also, take a look at /etc/rc.d/rc.local for missing <<EOL at the end. If any of those work, post here and I'll rephrase as an answer. I was struggling with those - but I think I nailed it now... –  Felix Aug 26 '13 at 5:50
add comment

1 Answer

If you mix EBS and instance store volumes, instance store data will be lost after reboot. For long term instances you should go 100% of times for EBS volumes, and leave instance storage class for "burst" applications that you only require to be online on demand; ASGs are the typical use case for instance storage.

The other possibility could be that you or your configuration management system mangled in some way your fstab file. If AWS volumes are not mounted in exactly the same path as detailed in the fstab file, instance boot will fail miserably. Also, do not forget that in EC2 boot volumes must be mounted to /dev/sda1.

share|improve this answer
add comment

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.