Re-installing a seriously ill Linode
As I found, the Linode control panel has some useful tools to help:
1) Using the control panel I shutdown the Linode and resized the disk to give myself 500mb of free space available.
2) In this 500mb of spare disk space I created an ext3 disk image, under 'configuration profiles' I then assigned this to become '/dev/ubdc'.
3) Booting up the Linode and logging into the shell prompt via the host console I mounted this (mount /dev/ubdc /mnt)
4) I then copied everything that needed to be retrieved to /mnt - the second disk image
5) Finally I shut down the Linode, removed the configuration profile (leaving the second disk image), then created a new configuration profile from scratch.
6) Mounting /dev/ubdc into the new Linode I then copied back the data.
I've still no idea why the original Linode went wrong, though suffice to say it was much faster to re-install than to fix it. Hope these pointers help somebody else in the same predicament.
Of course, there is no substitute for having off-site backups!
2 Replies
-Chris
@caker:
Sounds to me like you had your initial Linode to boot into Single user mode, and or booted with the "init=/bin/bash" option.
-Chris
Yes, I found this button in the control panel once I'd re-installed. I've no idea why it suddenly reverted to this mode though. Still, re-install worked wonders, apt is now working properly again and the machine is much faster.
I thought my experiences of copying the data onto a second disk might help somebody else out, hence the post.
Thanks Chris