[smartmontools-support] Lowest Fixed Version? Long test never finishes on WDC WD40NMZW-11GX6S1

Christian Franke Christian.Franke at t-online.de
Fri Mar 2 07:31:24 CET 2018


ItAint MeBabe wrote:
> I'm using an old version of smartctl ("smartctl 6.2 2013-07-26 r3841
> [x86_64-linux-3.13.0-142-generic] (local build)"  - this is what's
> current on Ubuntu 14.04)
>
> I am wondering if there is a newer version that addresses this issue,
> and if so, what version this was fixed in?  I'm hoping to find that on
> a Live Linux disk somewhere.

This is a drive firmware issue and cannot be fixed by smartmontools.


> I see this model,  WDC WD40NMZW-11GX6S1 listed in the latest
> drivedb.h, line 3865.

drivedb entries do not influence self-tests.


> This is similar to issue reported in:
>
> https://listi.jpberlin.de/pipermail/smartmontools-support/2017-December/000064.html
>
> The short test works fine in about 2 minutes.  The long test always
> says it's 90% remaining, and although the estimated time is 43
> minutes, the test still has 90% remaining after 8 hours.

Is there any regular I/O done during the self-test? If yes, please read:
https://www.smartmontools.org/wiki/FAQ#Whydoself-teststakeverylong


> (The WDC Windows tool estimates 160 hours - but I believe that it is
> *not* doing the actual S.M.ART test, because it is continuously
> reading 7MB/s from the drive.  It might be qucker if I wasn't using it
> thorugh USB 2 in a Windows VM, but as I said, it's not even a SMART
> test.)

Does a new test appear in the SMART self-test log after the WDC tool 
finishes the test?
If no, the tool did the test with regular OS reads.


BTW: If SMART self-tests do not work for some reason, I use something like

# ddrescue --ask -f -d --max-read-rate=1Mi /dev/sdX /dev/null disk.log

to perform read tests with limited bandwidth under Linux, Windows, ...

Thanks,
Christian




More information about the Smartmontools-support mailing list