I'm trying to recover a complex RAID that ran on my ESXi host. I'm using an Adaptec ASR-3805 Raid controller and was running two RAID 5EE on 8 x 500 GB drives. The first raid was 1126 GB (187.6 GB on each disk) and second 1664 GB (277.3 GB on each disk). I had two disks fail, which would corrupt the raid completely but was able to image half of one of the failed disks which should allow for recovery of the first raid using regions. I have the correct order, strip size (block size), and am pretty sure of the offset. I have also worked on finding the block layout and believe this to be the correct layout: http://www.redbooks.ibm.com/redbooks/SG ... =07-23.htm.
Now VMFS is not a supported file system for R-Studio, so I was trying to just make an image of the new Virtual Raid that is created and using vmfs-tools to read the image but it fails to do so. When I view the Virtual Raid device, I see empty data where the one missing disk is, so my question is when does R-Studio read the parity data and recreate the missing data? Is it only when Scanning the device? If so, is there anyway to have R-Studio read the parity data when creating an image of the device? I'm completely open to suggestions here, so any ideas?
A little more background, the only thing I'm interested in recovering is the data from my file server. It was an Ubuntu 8.10 Server with either a Ext2 or Ext3 partition. When running a Scan on the Virtual Raid device, it fails to read / recover larger files. I had originally thought this to be because of some incorrect raid parameter but now believe it to possibly be because of drive fragmentation as R-Studio does not understand the VMFS file system.