Seagate Barracuda 7200.11 ST31500341AS fw SD17
Markus
I have some Seagate Barracuda 7200.11 ST31500341AS 1.5TB drives. They
use the firmware SD17.
They time out at the flush write cache and sometimes on the identify
command.
I know that sounds suspicious so I wrote to seagate but they told me my
drives are ok. (Although I dont think so.)
This drive+fw is blacklisted in the kernel, but there is no firmware
upgrade for these drives. (Only very few serial numbers are able to
upgrade.)
Now I want to know after what period of the time the kernel timeouts at
these commands!? And if this timeout can be increased (manually, small
patch for the kernel) and the min/max/avg time could somehow be
messured. Every timeout kicks the drives off the raid, so I disabled
the write cache (hdparm -W0 /dev/sdx) but thats no solution, just a bad
workaround.
Its a amd64 system with the 2.6.29.4 on an AM3 board with AMDs 790GX +
SB750 Chipsets running in ahci mode.
Thanks for your time and support!
Markus
PS: Please CC me!
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Kay Diederichs
> Rui Santos wrote:
>>> I have some Seagate Barracuda 7200.11 ST31500341AS 1.5TB drives. They
>>> use the firmware SD17.
>> However, I do not know if it will solve your problem. Jut try it...
>
>>> This drive+fw is blacklisted in the kernel, but there is no firmware
>>> upgrade for these drives. (Only very few serial numbers are able to
>>> upgrade.)
>
> to "not affected serial number". I tried SD1B, the very newest version.
> Cant find the SD1A for my drives.
>
> Markus
I think you might have tried the wrong firmware file.
ST31500341AS needs
http://www.seagate.com/staticfiles/support/downloads/firmware/Brinks-4D8H-SD1B.ISO
HTH,
Kay
Markus
> Markus schrieb:
> > Rui Santos wrote:
> >>> I have some Seagate Barracuda 7200.11 ST31500341AS 1.5TB drives.
They
> >>> use the firmware SD17.
> >> Seagate recommends firmware SD1A for that HD.
> >> However, I do not know if it will solve your problem. Jut try it...
> >
> >>> This drive+fw is blacklisted in the kernel, but there is no
firmware
> >>> upgrade for these drives. (Only very few serial numbers are able
to
> >>> upgrade.)
> >
> > I cant upgrade the firmware because the flash-utility will abort due
> > to "not affected serial number". I tried SD1B, the very newest
version.
> > Cant find the SD1A for my drives.
> >
> > Markus
>
> I think you might have tried the wrong firmware file.
> ST31500341AS needs
>
http://www.seagate.com/staticfiles/support/downloads/firmware/Brinks-4D8H-SD1B.ISO
I tried that one (compared md5sum). It does not work. It only works
for "ST31500341AS, 9JU138-300 or -336, SD15 through SD1A ONLY!"
So the utility fails and says the specific model could not be found.
Markus
Krzysztof Halasa
> http://www.seagate.com/staticfiles/support/downloads/firmware/Brinks-4D8H-SD1B.ISO
>
> I tried that one (compared md5sum). It does not work. It only works
> for "ST31500341AS, 9JU138-300 or -336, SD15 through SD1A ONLY!"
Did you try to contact them?
--
Krzysztof Halasa
Robert Hancock
> Hello!
>
> I have some Seagate Barracuda 7200.11 ST31500341AS 1.5TB drives. They
> use the firmware SD17.
> They time out at the flush write cache and sometimes on the identify
> command.
> I know that sounds suspicious so I wrote to seagate but they told me my
> drives are ok. (Although I dont think so.)
>
> This drive+fw is blacklisted in the kernel, but there is no firmware
> upgrade for these drives. (Only very few serial numbers are able to
> upgrade.)
>
> Now I want to know after what period of the time the kernel timeouts at
> these commands!? And if this timeout can be increased (manually, small
> patch for the kernel) and the min/max/avg time could somehow be
> messured. Every timeout kicks the drives off the raid, so I disabled
> the write cache (hdparm -W0 /dev/sdx) but thats no solution, just a bad
> workaround.
I believe the timeout for commands like these is usually 30 seconds.
Normally if it doesn't respond after that length of time, it never will.
Even if it eventually did, I don't think you'd be happy with disk access
delays of over 30 seconds..
Markus
> Markus <M4rk...@web.de> writes:
>
> >
http://www.seagate.com/staticfiles/support/downloads/firmware/Brinks-4D8H-SD1B.ISO
> >
> > I tried that one (compared md5sum). It does not work. It only works
> > for "ST31500341AS, 9JU138-300 or -336, SD15 through SD1A ONLY!"
>
> Did you try to contact them?
Of course I did. They told me, my drives are not affected and drives
with "good" serial numbers will not be upgraded.
And they now told me I have OEM drives (???) which they do not provide
firmware upgrades for, although my drives are ok. I dont understand
those support-people.
Markus
Markus
> Markus wrote:
> > Now I want to know after what period of the time the kernel timeouts
at
> > these commands!? And if this timeout can be increased (manually,
small
> > patch for the kernel) and the min/max/avg time could somehow be
> > messured. Every timeout kicks the drives off the raid, so I disabled
> > the write cache (hdparm -W0 /dev/sdx) but thats no solution, just a
bad
> > workaround.
>
> I believe the timeout for commands like these is usually 30 seconds.
> Normally if it doesn't respond after that length of time, it never
will.
> Even if it eventually did, I don't think you'd be happy with disk
access
> delays of over 30 seconds..
Ok thats definitly bad. Will reask at the support...
Markus
Krzysztof Halasa
>> > I tried that one (compared md5sum). It does not work. It only works
>> > for "ST31500341AS, 9JU138-300 or -336, SD15 through SD1A ONLY!"
>
> Of course I did. They told me, my drives are not affected and drives
> with "good" serial numbers will not be upgraded.
> And they now told me I have OEM drives (???) which they do not provide
> firmware upgrades for, although my drives are ok. I dont understand
> those support-people.
Well, they might be right. Or the drives may be affected but they don't
support OEM versions anyway and may not know.
Obviously one could try updating anyway (I'm sure a simple hack in
version/SN checking routine would do). It may be risky.
You should, in theory, contact the OEM partner, not the manufacturer.
Could be hard.
OTOH I think the update fixes unrelated problem (a drives failing to
start), that would need checking.
--
Krzysztof Halasa