Just ordered some more drives to add to my array and have been reading a bit about this issue with hardware RAIDs, but some say even software RAIDs like unRAID are affected.  Is this true?  Should I be applying the non-supported FW that removes the timeout or am I safe.  I know I have another of these drives already in my system and then some slightly older 32mb cache 2TB WD drives as well.  All my drives are WD and reading about this now has me a little concerned about my data, so I am curious if I am worrying for nothing or is this truly something I need to correct before installing the drives and putting them into service.

 

thanks

It is irrelevant to unRAID.  TLER would allow the unRAID recovery for a sector to be invoked faster, rather than spend potentially several seconds with the drive doing its own error recovery.  This only comes into play when dealing with a bad sector, so in normal instances you will never see it.  TLER is intended for heavy I/O environments when waiting for the drive error recovery can affect the availability of the system to do other things.... i.e. when you don't want a bad sector to cause the array to bog down for several seconds.  Some HW RAID systems will timeout the drive before the drive's error recovery finishes, which is why TLER was created, so the drive will report the error, and stop trying to recover it, before the RAID times out the drive.  In a desktop (non-RAID) system, you lose data if the drive can't recover it, so you usually want the drive to try harder and longer.  But with RAID (SW or HW) you don't lose data if the drive stops retrying a bad sector.

Well thats good news.  I start reading things like that and get worried, but now I can rest easy once more.