Matt Miller wrote:
Try to run mdadm --Ebsc partitions. All though I doubt you can from the busybox. If nothing else changed, (even the OS is the same) just the two HDs and now you can't boot, I suspect you have a corrupt meta data issue on the raid. Now I am just guessing: use a live cd to see if the data is accessible (you will need one that speaks mdadm)? Are the two machines the same? Perhaps you are missing a required controller? A sure way is to re-install to another disk, have these two disks also installed in the box: when it comes to partitions choose your / and such, then pick raid, after it rights the changes to the disks and drops you into the raid set up, don't delete a raid, don't make a raid, just quit. then you will see your old raid pair. mount it at some data point, make sure NOT to format it. Boot to the new os, and collect your data.An etch machine died (machine won't power on), so I pulled the two hard drives out and dropped them into some identical hardware. These were the only two disks in the machine, and they formed a RAID1 set. Now the new machine can't find the root filesystem: Success: Loaded module raid1. Done. Begin: Executing udevtrigger again ... Done. Done. Begin: Waiting for root filesystem... ... Done. Checking root= bootarg cat /proc/cmdline or missing modules, devices: cat /proc/modules ls /dev ALERT! /dev/md0 does not exist. Dropping to a shell! Then I get busybox. How do I get this RAID1 set to work in the new machine?
There might be/is some way to scan the hd/partitions to register the array pair, but I don't remember what it is and you would still need an OS for it to work.
HTH -- Damon L. Chesser damon@damtek.com http://www.linkedin.com/in/dchesser