Whoops, broke the server 13/3/16

Hi all, sorry for the outage over Sunday evening and Monday morning.
I needed to power the machine down to move a power cable, so did some maintenance to remove old kernels filling up /boot and updated some packages. I didn't check at the end for the latest installed kernel. It was missing!
So after reboot it didn't come back up.

Installing the kernel again turned out to be a nasty crash course in reassembling RAID, mounting the partitions in the right place, enabling networking, chrooting and installing the kernel. Urrrghgh.

Then when installing the kernel, one of the drives suffered an error resulting in a RAID rebuild while in recovery mode which had to be left to complete before rebooting.

The server is back up now
But there is a problem to be tracked down with one of the drives. Its SMART status is OK, no reallocated sector count. So could be SATA cable or the drive interface electronics.

I expect the easiest way to solve the problem is to get an identical drive and swap out. The drive is £80, so any donations would be very welcome at this point.

Thanks for the info Thimo.
There haven't been any more errors while the server has been running. I wonder if the problem is just the cable or contacts /socket/soldering through the HD backplane.
Time will tell.