Dear Sebastien / Alain,
We have a client that had about 20 TB of Dicom files stored on a RAID5 array.
One disk failed (#16), but after replacing failed drive, during reconstruction, a 2nd HDD #8) failed too (causing RAID to go from degraded to failed).We have had sisk failure in the past, but never 2 disks of same array at same time…
After restart, the disk #8 came back online, and we started to replicate (rsync) the file to another server. Unfortunately, not only the copy speed was very slow (due to degraded array), but after a couple of hours the drive #8 failed again,
We achieved to copy 4TB of data, after several restarts, but now disk #8 looks dead.
I do not understand exactly how data is “spread” over the directories, but:
-
Is there a way to have (new install of) OrthanC “discover” the stored dicoms (and add them to the new OrthanC instance)? Using ImportDicomFiles.py?
-
If a study si “spread” over various directories, and part of the data is missing, will it import the “available” images, not import nothing, or cause failure in the import process?
We have backup (in offline disks) to reconstruct most of the 20 TB data, so we hope to be able to recover most data. But we will definitely migrate to RAID6, and we are now monitoring SMART HDD data to detect pre failure signs.
Many thanks in advance for you help.
Cedric.