[smartmontools-support] IBM/HGST/Toshiba(?) Super SMART

Christian Franke Christian.Franke at t-online.de
Mon May 21 19:28:34 CEST 2018


Игорь via Smartmontools-support wrote:
> There's a program, intended mostly for disk surface testing, named "Victoria for Windows v4.46b". It works correctly only on WinXP 32-bit.
> This program can also read SMART attributes, start drive self-tests, set AAM, etc.
>
> When reading SMART from IBM/HGST HDDs, it have an option, called "Use IBM Super SMART".
> When this option is turned on, program shows more detailed SMART information (more vendor-specific attributes in list).
>
> Moreover, there's one important fact, which is true for all other SMART reading programs I've seen so far. It's about SMART attribute 05, "Reallocated Sectors Count".
> If an IBM/HGST HDD have some reallocated sectors, any program I've seen, except above mentioned "Victoria for Windows v4.46b" (with "IBM Super SMART" option enabled), will be unable to see them, and will show raw value of SMART attribute 05 equal to zero.

Interesting. Thanks for the info.


> Which leads to "funny" situations, when, on IBM/HGST drives, raw value of SMART attribute C4 ("Reallocation Events Count") is some tens or even some hundreds, and, at the same time, raw value of SMART attribute 05 ("Reallocated Sectors Count"), is shown as 0. Despite the fact that, in fact, it also amounts to several tens or hundreds.
>
> Therefore, here's a question: can "IBM Super Smart" support be implemented in smartctl?

There is apparently a vendor specific SMART command (FR=0xf8) to 
en/disable "Super Smart" mode:
https://forum.hddguru.com/viewtopic.php?f=7&t=35736&start=25

Is there any official specification available?

Feel free to create an enhancement request in our trac:
https://www.smartmontools.org/newticket


> Or, at least, correct display of SMART attribute 05 for IBM/HGST drives?

No. Smartctl simply prints what the drive returns in its SMART data 
block. If it always returns 0 for this attribute, there is nothing we 
could do.


> BTW, I have a strong suspicion, that this also applies to HGST/Toshiba 3.5" drives, and, maybe, HGST/Toshiba routines can be tryed with IBM/HGST drives.

How could be checked whether a drive supports this command? Issuing 
vendor specific commands to drives which do not support it may be dangerous.


Thanks,
Christian

PS: For future submission, please avoid screenshots if possible. Mails 
with binary attachments may be delayed or blocked by the mailing list.




More information about the Smartmontools-support mailing list