A follow on from this story and many moons after my "fickle" switch to Fedora/RHEL.
At this point in my life I'd been involved in a half dozen large companies and used linux on enormous scale.
I moved to a company that was using ubuntu LTS (10.04) (old at the time) in production, it was heavily invested and I expected that wouldn't change as Developers were very hesitant to change to debian (which is too old/doesn't make things easy enough) or centos/RHEL which suffers the same issues and has the added benefit of having SELinux (which I'm an advocate of understanding rather than disabling).
I go through my daily security advisories and a local privilege escalation means all our virtual machines and virtual machine hosts are affected, luckily it's patched as 10.04 is still supported so I apt-get update;apt-get upgrade and send out an email saying the server will be down for 30 minutes while it receives patches.
I was wrong, it was down for 6 hours.
unfortunately someone upsteam caused that particular kernel update to rebuild all initramfs' on the machine, and had also named lvm2 to lvm, so now my drives wouldn't mount.
On any kernel version/initramfs version
normally you can drop to shell load the module, mount the drives and continue startup, but unfortunately that stopped a lot of things from loading such as the bonding we had in place on the nics.
obviously I didn't know why it broke at the time and was attempting to get help from #ubuntu on freenode.
the response was "Sometimes it's better not to know why it broke"
that server was smoothly running CentOS before I left that company.
So in my opinion support and enterprise is where it falls down.
At this point in my life I'd been involved in a half dozen large companies and used linux on enormous scale.
I moved to a company that was using ubuntu LTS (10.04) (old at the time) in production, it was heavily invested and I expected that wouldn't change as Developers were very hesitant to change to debian (which is too old/doesn't make things easy enough) or centos/RHEL which suffers the same issues and has the added benefit of having SELinux (which I'm an advocate of understanding rather than disabling).
I go through my daily security advisories and a local privilege escalation means all our virtual machines and virtual machine hosts are affected, luckily it's patched as 10.04 is still supported so I apt-get update;apt-get upgrade and send out an email saying the server will be down for 30 minutes while it receives patches.
I was wrong, it was down for 6 hours.
unfortunately someone upsteam caused that particular kernel update to rebuild all initramfs' on the machine, and had also named lvm2 to lvm, so now my drives wouldn't mount.
On any kernel version/initramfs version
normally you can drop to shell load the module, mount the drives and continue startup, but unfortunately that stopped a lot of things from loading such as the bonding we had in place on the nics.
obviously I didn't know why it broke at the time and was attempting to get help from #ubuntu on freenode.
the response was "Sometimes it's better not to know why it broke"
that server was smoothly running CentOS before I left that company.
So in my opinion support and enterprise is where it falls down.