Recently it was my displeasure to work on a three drive spanned set for NT 4.0. The set was soft configured so when the boot drive went down, with all of the configuration data for the RAID on it, the RAID would not mount. In addition to that, this was a multi volume set. In other words, there were two spanned sets within the three drives, staggered across all three drives.
Two problems existed with the RAID. First, the spanned set, like most spanned sets had their starting points staggered. Second, each set was staggered across the three drives. If you are familiar with DELLs channelling system for configuring multiple physical devices using portions of each drive then you are familiar with this RAID handling scheme. For those who are not familiar, here is how it works.
As an example lets take three drives. Each drive is 36 gigabytes in size. For a spanned set you would normally take all of drive one, all of drive two, and all of drive three. These drives would then be considered contiguous in the array. In other words, you fill up drive one first, then drive two, then then last drive. With this schema you can usually tell when you have a spanned set by mounting the first drive and taking a piece of data recovery software and treeing the drive up. Since the Master File Table is at the beginning of a partition the recovery software will tree the drive as there are no drive seek and reads past the end of the first drive. You can actually retrieve data from the first drive if all of the data resides on the first partition. However, any data that is stored on either drive two or drive three will not be retrieved as you have only the first drive mounted.
This configuration sounds a bit complicated. That’s because it is. It can be very difficult to decipher start and end points for the spanned sets where the entire dive is used for the span. Now, add this little wrinkle. The configuration not only has one spanned set, it has two. So now you have six points of origin, six span chunk sizes, and two drive partitions with valid data. What a mess.
Next time I will have real numbers to use from a set I recovered and show you how you can recover your RAID with no RAID recovery software.
Until next time…
We have the following problem: Our server (propably HP) have hardware RAID 5
configuration. Customer informed us that during power outage the server crashed and later after rebooting the D: drive is missing. What do you think can be a case? There are all information what I have right now.
Thank you for your help.
Chris Montwill