Two SMART properties check programs, different results of End-to-End error

Question

Take a look at SMART-ID value D8:

Which one of the two is correct and is the current state really critical? The RAW values are NOT different:

• Crystal reads 11h = 17
• Passmark reads 17 = 11h

Answer 1

To get you in the right direction because your question shows minimal effort and research.

Technically normalized value is below threshold so in simplistic view of tool on the right a SMART attribute failed. Below threshold = FAIL for the particular attribute.

So rather than quibble about what tool is right, look into end-to-end-error or Reported IOEDC errors attribute.

Seagate's description:

Attribute ID 184: Reported IOEDC errors Normalized Reported IOEDC Error = 100 – Life Time IOEDC Errors. This attribute tracks the number of IOEDC errors encountered during host initiated reads and writes.

Raw Usage

Raw [3 – 0] = LifeTime IOEDC Count

It sort of detects and counts errors that occur while data is being transported from host to disk. So AFAIK it can be cable 'noise' as well as RAM buffer errors.

In Seagate's own words:

If IOEDC is triggered during the transfer from FIFO to disc, the data will be re-transferred from buffer up to disc 4 times. If IOEDC still fails the drive will write uncorrectable data to Media. When a subsequent read is issued to this located a hard error (uncorrectable) will be returned to the host, not unnecessarily failing IOEDC, causing a host level retry.

I'd back up data and run extended SMART test and go from there. Other than that you can check the drive using the manufacturer's utility (Seatools).