On 9/13/26 17:40, Gary Dale wrote:
Your advice would work in many situations but I think you missed that
the drive is part of two RAID-6 arrays, only one of which is degraded.
Also it is showing up in the device listings, I can used gdisk on it, it
passes SMART testing. There is nothing significant to show a drive
problem. And a power issue should show up in more than one drive.
I can't fsck the partition obviously since it doesn't have a file system
- just the mdadm superblock, if zeroing it left anything.
I think the problem must have something to do with mdadm but I'm not
sure what. Being RAID-6, I'm not overly worried at this point as it
still has one redundant drive. However I am using RAID-6 for a reason.
Unfortunately, I have discovered that my reasoning can be wrong and that
I must embrace Einstein's definition of insanity. I have found that
starting at the bottom and working my way up the dependency/ complexity
tree gives me the best chance of success.
AIUI you have four disks with two partitions each, you have put all of
the first partitions into one array, and all of the second partitions
into a second array. I would expect such a configuration to cause disk
contention whenever the two arrays are in use at the same time. I am
curious -- why did you not take the conventional approach of one array
using one large partition per disk, then subdivide with LVM?
David
