I've got a 4 drive RAID5 array. One of the drives quit working quite some time ago. Another drive (/dev/sdc)just started developing read errors so that now the array will not mount.

I've got a new drive (same model# as the ones in the RAID5 - 250 gb) and want to copy the data from /dev/sdc to the new drive. I've got ddrescue ready to execute. Running Ubuntu Server 8.10 with the OS on /dev/sdd - a dedicated 80 gb hdd.

• Can I disconnect the other drives in the RAID5 so that the only one connected is /dev/sdc (the source for the copy)?
• Can I simply connect the new drive to one of the newly emptied slots and run ddrescue?
• Or, do I have to issue some kind of commands to do a low-level format and designate the drive something unused (such as /dev/sdf)?
• What is the process for installing the virgin hdd and copying the data to it?

Thanks,
David

I didn't see this when you 1st posted it. I may be able to help, but 1st tell me if you solved this already.

Rick,

Thank you for the response.

Short answer is "Yes, it has been resolved."

Longer answer is that I found the San Antonio Texas Linux User's Group (SATLUG) here in town, which I had joined their email list but never really viewed because it seemed more social or different from what I was needing at the time. However, I posted my problem to their email membership, and received an amazing amount of hand-holding that get me recovered. They helped me

• Determine which disk was bad and which one just had old data (more on this below),
• Walked me through ddrescue to extract the data off the disk that had gone bad and get it onto a new disk,
• Worked through rebuilding my array with the three current disks (two good ones & one new one with data copied from the bad one),
• To get the old one (that had not been used in a while - apparently coming loose from the data cable and I had not noticed/checked it) working again,
• To get the 4 drive array rebuilt,
• Find/fix/remove any files that contained errors,
• & finally, be up and running.

The whole process took about 2 weeks, but there were two guys in the group who were absolutely amazing, even exchanging emails back and forth until the wee hours of the mornings as I was working through the processes.

Amazingly enough, out of 200+ GB data in over 35,000 files, I had only about 6 pictures that did not recover!

I have since installed a 1tb HDD in my desktop and keep a complete backkup of the data there, swapping out the backup drive occasionally for a blank one - rotating them grandfather-father-son style and keeping the latest ones off-site while only having the current grandfather on-site to be ready to do the next backup as data is accumulated. I've also been routinely running disk check routines to make sure everything continues to function properly. I've also got a UPS installed, but have NOT yet installed the auto-shutdown software to insure an extended outtage won't shut down the system unexpectedly. What I have NOT yet done is configure the server to email me when something starts to fail. I need to find the steps to make that happen.

And, I really do appreciate your checking up on me. Thank you.

If you have any tips on the notification thing, someplace I can look for info, that would be WONDERFUL!!!