SUCCESS!! I was able to restore the data thanks to the WIKI and your emails. Thank you so much! Now, how did I do it? When first attempting to re-create the RAID I knew that the disk order, the number of disks, the RAID level and the style of the superblock were important, but I didn't take care of the chunk size. Yesterday, I learned from you guys that the defaults that mdadm --create uses changed over time and that the chunk size was one of them. Comparing the mdadm --detail output from the re-created RAID on my cloned drives with the output obtained from the original RAID showed me that it was 512K vs. 64K on the original RAID. Obviously, the clones were not junk yet, as adding a --chunk=64K to my initial mdadm --create command was successful in my case. I then just had to do a vgchange -ay and boom, there it was, my volume showed the correct size and mounted like a charm. 100% data secured. Anyhow, I learned a lot during the process; first of all that even a 2-disk failure tolerance on a RAID 6 is not a reason to get lazy with the backups. But also I learned a bit of how a RAID actually works and what to do when it fails. Thanks again and have a great day/morning/evening, everybody. Really appreciate your immediate support! Best regards, CygX1 On 24/10/18 5:14 pm, Wols Lists wrote: > On 23/10/18 23:56, CygX1 wrote: >> Yes, creating the copies takes 8 days via USB 3.0 :-( -> that's actually >> the reason I copied just two drives and not the complete set of 4. > Which is why "replacing a failed drive", on the wiki, recommends in the > first paragraph that you get yourself an add-in SATA card. A cheap > 2-port card should be about £30. (Tech exchange rates tend to be £1 == $1.) > > You're probably looking at about £50-£100 for a cheap data recovery > setup - add-in card and cradle - that'll make disaster recovery for > yourself or others much easier. And I've made a few quid on the side > recovering data for people who knew I had the ability ... :-) > > Cheers, > Wol
