clear smart error log on wd red 8tb

[tofurky]

hi,
i recently purchased several 8tb wd easystore usb3 external drives which contain WD80EFAX-68LHPN0 wd red drives. during migration of my data, i had a poor esata connection to one of the drives and encountered a few UDMA_CRC_Error_Count, which also put errors in the log. because the drives are brand new, and the error wasn't due to the drive, i was hoping to clear out the few crc errors before putting the drive in service in my nas.

i purchased a 1 month license of wdmarvel, but apparently even though it's a "wdc" drive, it is actually a rebranded hgst internally. wdmarvel didn't detect it, but did work on older wd drives (manufactured in 2003 and 2013) i had. after emailing their support they agreed that it was actually an hgst (helioseal i guess?). the firmware is "nasware 3.0" 83.H0A83.

since wdmarvel couldn't detect it, i tried the sediv trial, but it doesn't even let me select hitachi from the menu. i don't see a way to clear smart using the demo either.

my question is, are there any utilities or methods that will let me clear the smart log of this drive that are free or cheap? the drive is technically fine, but it bugs me to see errors in the log via smartctl.

thanks,
-matt

[tofurky]

hi,
just a small update. i was able to try wdr and sediv after changing my controller from ahci to ide. still, the drive didn't seem to be detected as a wd and the "clear smart" commands had no effect.
thanks,
-matt

[BGman]

my question is, are there any utilities or methods that will let me clear the smart log of this drive that are free or cheap?

thanks,
-matt

Yes, but nobody is going to tell you!

[tofurky]

Yes, but nobody is going to tell you!

lol why not? i promise i don't work for the competition since i'm new my other reply didn't go through (yet?) but i did change the controller from ahci->ide and got sediv and wdr to detect the drive, but none of the commands seemed to do anything

[BGman]

lol why not? (

They may tell you "send it to me and I'll fix it", but nobody is going to tell you how!

Trust me.

[labtech]

@tofurky

What NAS model do you have?

Thanks.

[tofurky]

@tofurky

What NAS model do you have?

Thanks.

hi labtech,

it's a custom built one using a supermicro x7spa-h motherboard with a cfi a7879 case/backplane. the esata card is mediasonic hp1-ss3. filesystems are btrfs, os is debian stretch.

the esata enclosure/cable/card all seem fine. i believe it may have been a bit of dust that made its way into the connector while i was rearranging/installing the new drives. after a bit of compressed air and a reseat of the connector, i was able to read several tb from the disks (raid0 for the esata ones) without encountering further errors. i bought some new shorter, flexible esata cables that i plan on using to hopefully ensure a better connection in the future.

since it runs only linux, i have been hooking the drive up to my intel z77 based workstation via esata and booting into windows to run the various utilities (wdmarvel, wdr, sediv, victoria, etc.)

thanks,
-matt

[tofurky]

thanks all for the replies.

i have done some more digging and come across some threads in russian forums (i can't understand russian...) where 2 utilities are discussed.

one is called FormUnit from what i gather, and the other Hitachi_format or HGST_fmt.exe. i haven't tried either yet, but it seems they clear at least some smart attributes (but not power on hours, for example). no telling if they'd work on a ibm-hitachi-wd

i'd also prefer someone do it remotely, if possible. if the FormUnit command doesn't work, is there anyone who'd do a remote session (e.g. teamviewer or chrome remote desktop) to BartPE or similar and execute the commands that way? shipping the drive introduces the rough handling of the courier, and i could buy 2 new drives for $350 lol.

thanks,
-matt

[BGman]

FormatUnit will erase all your data and will clear only Atr. 5 of the SMART data.
It takes me less than a minute to clear SMART of all Hitachi drives and it's always free for my clients.

[tofurky]

someone outside of here was kind enough to send me an mhdd script which executes the hgst clear smart vsc. i haven't tested it yet but will do later.

thanks,
-matt

[fzabkar]

I suspect that the UDMA CRC Error Count may be one of those attributes which compute a rolling average rather than a cumulative count over the drive's lifetime. If so, then the attribute value should clear up of its own accord, given sufficient time.

[fzabkar]

I'd like to understand the "logic" behind pricing an internal 8TB drive at $50 - $100 more than the same drive stuffed into a USB enclosure with a DC adapter and USB cable.

[BGman]

Of course one still needs a "valid" Super ON for the specific drive. Super ON for IBM/HGST/Hitachi change very often so one needs to either compute the correct Super On or use a table of Super On commands for the specific drive.

Are you sure?

It doesn't need "techno mode", as far as I know.

[tofurky]

hi,
from what i can tell, the "super on" command has been the same probably since at least 2008. i want to say i found a much earlier reference to it but i can't find it in my history

from the victoria changelog from 2008-08-13 4.46b: "Added support for super-SMART for IBM / HITACHI ("factory" attributes)"
from http://forum.ixbt.com/topic.cgi?id=11:42461-72, there is a link to a tool which sends the command, ibm_ss.exe: http://vlo.name:3000/tmph/ibm_ss.rar

i attached an image showing the disassembly of ibm_ss where it sends the super on/super off - the command is clearly visible and the same.

i can't find any references to the clear smart vsc, however. just the super on/super off that *seems* to go back to ibm drive days.

the mhdd script i obtained did clear many smart attributes from the WD80EFAX drive, but it unfortunately did not clear the actual error log. UDMA_CRC_Error_Count did go to 0. my main goal was to clear the *log* though Power_On_Hours and other wear-related attributes were not changed. i attached a before & after of the attributes cleared if anyone is interested.

if anyone knows the vsc or otherwise to clear the actual log, i'd be very much appreciative

and, as far as the logic of selling the same drive for $100 less as an external vs. internal, well. i guess western digital doesn't have any other (cheaper) 8tb skus. they have the purple but maybe it's not suitable for desktop/external use?

i did make a quick collage of an earlier gen wd red drive (WD20EFRX), my drive (WD80EFAX), and a hitachi helioseal 8tb (not sure of model). the WD80EFAX and Helioseal look identical, so that for me at least answered the question of it's a rebranded hitachi. this might be a "well duh" for the rest you although this is supposedly 5400 rpm while i only see 7200 rpm skus on the hitachi website...

thanks,
-matt

Attachments

beforeandafter.txt

(1.17 KiB) Downloaded 1642 times

supersmart.png (11.94 KiB) Viewed 30457 times

[BGman]

Quote:
"SMART Function Set Command (B0h)
The SMART Function Set command provides access to Attribute Values, Attribute Thresholds and other
low level subcommands that can be used for logging and reporting purposes and to accommodate special
user needs. The SMART Function Set command has several separate subcommands which are selectable
via the device‟s Features Register when the SMART Function Set command is issued by the host."

[fzabkar]

ISTM that only the performance related attributes are cleared.

Code:

                                 BEFORE:                 AFTER:
                                 ID;Current;Raw          ID;Current;Raw

Raw_Read_Error_Rate             1;100;0;                1;100;0;
*Throughput_Performance          2;127;120;              2;100;0;
*Spin_Up_Time                    3;161;34385559976;      3;100;0;
Start_Stop_Count                4;100;24;               4;100;24;
Reallocated_Sector_Ct           5;100;0;                5;100;0;
Seek_Error_Rate                 7;100;0;                7;100;0;
*Seek_Time_Performance           8;128;18;               8;100;0;
Power_On_Hours                  9;100;272;              9;100;272;
Spin_Retry_Count                10;100;0;               10;100;0;
Power_Cycle_Count               12;100;21;              12;100;21;
Helium_Level                    22;100;100;             22;100;100;
Power-Off_Retract_Count         192;100;32;             192;100;32;
Load_Cycle_Count                193;100;32;             193;100;32;
*Temperature_Celsius             194;253;188979806231;   194;240;115965755419;
Reallocated_Event_Count         196;100;0;              196;100;0;
Current_Pending_Sector          197;100;0;              197;100;0;
Offline_Uncorrectable           198;100;0;              198;100;0;
*UDMA_CRC_Error_Count            199;200;3;              199;200;0;

[tofurky]

ISTM that only the performance related attributes are cleared.

yep. according to the person who sent me the script, it *used* to clear more, but recent hgst (and "wd") don't clear the wear-related ones. which is good, imo. i think it shouldn't be that easy to wind back an "odometer"...

[BGman]

3 - Smart function set is used as mentioned but on this case it's VENDOR SPECIFIC. Meaning this set of commands will only work on Hitachi/IBM/HGST and will NOT work on other drives, so at the end of the day it's VSC even if it uses a "standard" command as basis.

I would say vendor specific subcommands of Smart B0h command.

Quote:
T13/1699-D Revision 4a
May 21, 2007
Working Draft AT Attachment 8 - ATA/ATAPI Command Set (ATA8-ACS)
7.53 SMART
Individual SMART commands are identified by the value placed in the Feature field
.......
E0h-FFh vendor specific

[tofurky]

so as a last ditch effort (and because i was kind of planning on doing a secure erase anyways) i spent some time trying to get igor_me's IBM/Hitachi format program (i.e. Format Unit) working on windows 10 64-bit. somehow, after disabling integrity checks and enabling test signing via bcdedit, downloading the WinIo .dll/.sys 64-bit driver separately, and installing a utility that allows specifying locale (e.g. russian) on a per-program basis, i got the utility working and was able to translate the various text in the program using ocr+google translate.

well, 10 minutes into the format of the WD80EFAX and i get a blue screen "driver power state failure" and my pc reboots. now, the drive is not recognized. it is hooked up via esata with an external power supply, so the power wasn't necessarily cycled on the drive during the blue screen. i suspect it may still be executing the format and ignoring ata commands until it finishes.

does that sound right or even possible? i'd rather not interrupt the format if it's still going, because it'll probably crash again if i restart it. but i also don't want to wait 20 hours to see if the drive is actually formatted.

[BGman]

FormatUnit will erase all your data and will clear only Atr. 5 of the SMART data.
It takes me less than a minute to clear SMART of all Hitachi drives and it's always free for my clients.

You were told what to expect from FormatUnit !
The problem is that the script you were given can clear only a few attributes. I tried these commands on a relatively new HGST drive and the result was as yours. Then I used my old utility and it cleared all attributes , including POH !