Increasing reallocated sector count but SMART says good...

[milks]

Hey


I swapped out one of the disks in my Ultra2 a couple of weeks ago because of the following warnings:

Detected increasing reallocated sector count[14762] on disk 2 [ST32000542AS, 5XW1F84E] after poweron -1 hours. This often indicates an impending failure. Please be prepared to replace this disk to maintain data redundancy.

but when I hook the drive up to my laptop and check the SMART info it reports a health status of good. Here's the full breakdown:

----------------------------------------------------------------------------
(2) ST32000542AS
----------------------------------------------------------------------------
Enclosure : ST320005 42AS USB Device (V=1631, P=3200, sa1) - st
Model : ST32000542AS
Firmware : CC38
Serial Number : 5XW1F84E
Disk Size : 2000.3 GB (8.4/137.4/2000.3/2000.3)
Buffer Size : Unknown
Queue Depth : 32
# of Sectors : 3907029168
Rotation Rate : 5900 RPM
Interface : USB (Serial ATA)
Major Version : ATA8-ACS
Minor Version : ATA8-ACS version 4
Transfer Mode : ---- | SATA/300
Power On Hours : 27540 hours
Power On Count : 31 count
Temperature : 20 C (68 F)
Health Status : Good
Features : S.M.A.R.T., APM, AAM, 48bit LBA, NCQ
APM Level : C0C0h [ON]
AAM Level : FEFEh [ON]

-- S.M.A.R.T. --------------------------------------------------------------
ID Cur Wor Thr RawValues(6) Attribute Name
01 119 _99 __6 00000C9B9794 Read Error Rate
03 100 100 __0 000000000000 Spin-Up Time
04 100 100 _20 000000000020 Start/Stop Count
05 100 100 _36 000000000000 Reallocated Sectors Count
07 _82 _60 _30 00000B4EB08A Seek Error Rate
09 _69 _69 __0 000000006B94 Power-On Hours
0A 100 100 _97 000000000000 Spin Retry Count
0C 100 100 _20 00000000001F Power Cycle Count
B7 100 100 __0 000000000000 Vendor Specific
B8 100 100 _99 000000000000 End-to-End Error
BB 100 100 __0 000000000000 Reported Uncorrectable Errors
BC 100 _99 __0 000000010006 Command Timeout
BD 100 100 __0 000000000000 High Fly Writes
BE _80 _49 _45 000014130014 Airflow Temperature
C2 _20 _51 __0 000D00000014 Temperature
C3 _36 _26 __0 00000C9B9794 Hardware ECC recovered
C5 100 100 __0 000000000000 Current Pending Sector Count
C6 100 100 __0 000000000000 Uncorrectable Sector Count
C7 200 200 __0 000000000000 UltraDMA CRC Error Count
F0 100 253 __0 1F8A00006BE0 Head Flying Hours
F1 100 253 __0 0000F9F9692F Total Host Writes
F2 100 253 __0 0000A5701E92 Total Host Reads

any chance the error was a false positive?


thanks 😎

[RX]

just want to check if you run the Extended test using the Seagate SeaTools on your disk model: ST32000542AS?

[StephenB]

I would definitely run seatools, and I'd try the destructive write test as well as the extended read.

The -1 power on hours is a bit confusing, so possibly there is a bug in the original report. But I think running full diags is a good precaution.

[milks]

thanks chaps. I'll run the tools tonight and report back

[sgogo]

I am never sure I know how to read the SMART data properly on Seagate drives, but I do notice what might be a high command timeout (000000010006) value.

SteveG

[StephenB]

Yes, I saw that as well. But I'm also not sure how to interpret the raw binary value.

[mdgm-ntgr]

Did you run SeaTools as suggested?

[milks]

hey

I wasn't too sure which test to run :roll: so I ran the one which sounded most thorough. It took a couple of hours but passed fine



I wasn't sure what the hex values on the SMART report meant, I thuoght they might be the memory address for each metric. I was looking mostly at the current and worst values in columns 2 and 3. Anyway 0x000000010006 command timeout is 655542 in decimal - still doesn't mean anything to me

[StephenB]

I'd also suggest the write-zeros test (which is destructive, but shouldn't matter). It's under advanced.

Your smart report is not reporting addresses, it is reporting values. Some vendors define bit-fields for some parameters(e.g, the "1" might mean something else). It's only 65542 btw (small typo in your post). It is plausible that it is a simple count.

Not sure if seatools reports the smart values or not - you might poke around and see if you can find them. There are other tools (acronis disk monitor is a free download, though you do need to leave your email address).

[sgogo]

I agree with StephenB.

The vendors have a proprietary implementation on most of the SMART values, so it is hard to tell if something is wrong by looking at the raw values (in either hex or decimal). The meanings even vary depending upon drive type and even firmware revision on similar drives. Only the vendor (Seagate) can properly understand what is going on there.

However, I have never had a drive fail the SeaTools testing...so I am not sure how much value that really is for predicting failure. I think it is really more of documentation issue to get a drive RMA'd within the warranty period.

My rule of thumb is to use the following on SMART values as a reference point... If any of these are going up, I dump the drive for critical storage:

-01 - Read Error Rate *
-05 - Reallocated Sectors
-BB - Uncorrectable Errors
-BC - Command Timeout
-BF - G Sense Error rate *
-C3 - Hardware ECC recovered *
-C5 - Current Sector Pending
-C6 - Uncorrectable sector count

For Seagates, the "01 - Read Error Rate" should equal the "C3 - Hardware ECC recovered".
If it doesn't, that does not mean its bad, just that you have to look deeper at the hex values in the first 4 bits to see how bad it is, and watch those 4 bits.

Also for Seagates, I do not know how to read the command timeout values. I generally see numbers there on all drives, so I cannot imagine that every one of my Seagates is bad... however yours are kind of high for the amount of hours on the drive. Possibly this means a bad cable and/or your ReadyNAS is doing something wrong.

For "BF - G Sense error rates", I only care if it is rising when the drive is reading/writing while not being moved on a steady surface.

BackBlaze has done studies on this (google it), but reading SMART is kind of like reading tea leaves, IMHO.

SteveG

PS - I use CrystalDiskInfo or HDD Sentinel free versions to check SMART values. Some versions of CrystalDiskInfo have adware, so do some research before installing.

[StephenB]

The threshold for failure in the vendor diagnostics is higher than my personal threshold. Seatools will certainly pass with more bad sectors than I am comfortable with, so you should review the SMART stats again.

Command Timeouts are a bit hard to sort out, but if they are increasing on one drive and you have the same drives in the NAS with no timeouts, then that is a clue that something is wrong.