After my downtime I chose to continue with using Amazon’s EC2 service as my primary host. Despite the one issue I had, the overall qualities of EC2 are ten times better than other hosts I’ve used.
One of the problems that several people have told me they had with EC2 relate to their use of out-dated software. The base images they give for starting up your server are all Fedora Core 4 (FC4) images. The problem with that is, FC4 is no longer supported.
(There are, of course, contributed images of Fedora Core 6, but the quality of them should be considered questionable given that you’re not sure who/where they came from.)
Since I already had some downtime, I decided it’d be worth it to have a little more and upgrade my server. There were quite a few snags in upgrading the FC4 image to FC6, so I thought I’d document exactly what I had to do to get it upgraded.
The first thing I did was start a new instance. If you don’t know how to do that, you shouldn’t be reading this post.
Next, the list of commands I used (as root):
- yum clean all (just for kicks)
- rpm -Uhv http://download.fedora.redhat.com/pub/fedora/linux/core/5/i386/os/Fedora/RPMS/fedora-release-5-5.noarch.rpm (to set the path for upgrading the kernel to the FC5)
- yum remove fuse-encfs\* (because that package will fail when upgrading)
- yum upgrade udev (because that package must be upgraded before the rest for some silly reason)
- yum install kernel (upgrade the kernel, part 1!)
- reboot
- yum clean all
- rpm -Uhv http://download.fedora.redhat.com/pub/fedora/linux/core/6/i386/os/Fedora/RPMS/fedora-release-6-4.noarch.rpm http://download.fedora.redhat.com/pub/fedora/linux/core/6/i386/os/Fedora/RPMS/fedora-release-notes-6-3.noarch.rpm (it’s FC6 time!)
- yum remove kernel-2.6.17 kernel-2.6.16 (two other kernels exist when you only want one; if you don’t remove these, the upgrade will fail)
- yum update (upgrade the kernel and everything)
- reboot
It really wasn’t too bad, overall. It just took some time to figure out the (weird) issues.
Note: Some of these steps may not actually be required, but these is the path that worked for me. Your mileage may vary. As always, try this on a random instance first before a “real/production” one. If it fails, you can just turn it off.
When you say boot up a new instance, do you mean a new instance of a customized instance, or a fresh base package? Also, how did you migrate your hosting data from one instance to the next? Did you do this live in the background?