It’s 6/6/06 today – birthday of the Beast 😉
I think one of Solo’s hard disks is on it’s way out. Not great as I’ve only had it a year. The problem is, it’s very hard to tell which one it is, as it’s very intermittent and as the two drives are LVM’d together, it’s hard to tell which disk holds any corrupted files.
At one point the computer booted without showing disk 1 (so maybe power problems?) however that’s the data drive, and I think it’s disk 0, the boot drive that’s the problem, as last time it died I continued working on the VMWare instances on the data drive….
It’s possible it’s also caused by a SATA_SIL bug in the Linux kernel, especially as WD’s diagnostic tool doesn’t show any errors – but that only reads the disks, no writes.
I bought them from Scan, so we’ll have to see how nice their RMA system is, maybe I can return both drives as they are a RAID array and were bought together – and they have a 3 year warranty from Western Digital.
Another worrying thing is that not only do I have these two WD2000JS’s in Solo, I also have two WD2500KS’s in Greivous, as does Kev!
Update: I’m 99% sure it is the first disk that’s the problem, as looking at the LVM config, for some weird reason I made disk 0 (sda) data and boot, and disk 1 (sdb) swap and root. So if disk 1 failed it would still boot like it did.
I’ve copied all my VM’s and useful stuff onto my external drive for backup, most of it’s on DVD as well.