On the weekend I had to come into work because our development server (which I won’t give the DNS of) started to have data coruption issues. We had recently uprgraded the drives from a pair of 40Gs to 120Gs. In this graduated process we started with just moving the information onto the new drives then from there running it off one drive for a week or so and then rebuilding the RAID1 system (two drives working as one, with one as an instant backup).
The RAID build gave a great failure at 100% error. I had assumed we could leave it for the weekend and give it another go on Monday. Apparently not. The data on it started corupting and the drives were throwing short.
So I came in, backed up what had changed since it’s tape backup. Swapped one of the drives for another of the same type/size. Reinstalled RedHat AS3 on it twice – still getting drive errors. FINALLY it occured to me, the 14′ drive cables we bought to facilitate the size of that case and it’s moving drive bay. Swaped the 14′s for 12′ – solved all the problems. Started the long process of restoring over 100MB network card.
But really, who ever hears or gives adivce: “14′ cables work well, but just don’t use two, your computer doesn’t have the power to cover that kind of distance” You hear about bad drives, boards, or even broken cables – but cables that are too long to be paired? That’s a new one for me.