AHCI / Intel Rapid Storage 9.6 - bug on Windows XP : drives don't go to sleep anymore.

[idata]

Hello,

I would like to report an issue to Intel: when using "Intel Rapid Storage" version 9.6 on Windows XP SP3 32bit, the drives will NOT go to sleep anymore.

If in Bios I set SATA back to "IDE" (Intel Rapid Storage isn't used anymore), then the drives will sleep perfectly.

The data:

The board is Asus P5K-VM with Intel ICH9 controller.

OS is Windows XP SP3. In Bios, SATA is set as AHCI.

The Intel Rapid Storage driver (iaStor.sys) installed version is 9.6.0.1014.

The HDDs are Samsung HD103UJ, there exist no firmware update for those drives.

In the XP "Power Options", I set "Turn off hard disks" to 3 hours, unfortunately, the drives will *never* go to sleep after 3 hours with AHCI/ iaStor.sys.

If in Bios I change SATA back to IDE (iaStor.sys isn't used anymore) then the drives will go to sleep after 3 hours...

I can reproduce the issue _every_ time, it's consistent.

There seems to be a small bug in iaStor.sys version 9.6 that prevents the HDD to go to sleep after 3 hours.

An interesting point, using AHCI/iaStor.sys and setting "Turn off hard disks" to 15 minutes DOES work. But set to 3 hours, it won't.

Using IDE, it will go to sleep always, 15 minutes, 1 hour, 3 hours, etc...

Thanks for your support,

TSR

[idata]

Following up on this bug, with more testing.

I discovered that HDDs will not go to sleep if the time set for "Turn off hard disks" is HIGHER than 45 minutes.

Anything below and up to 45 minutes included put the HDDs to sleep correctly.Anything higher than 45 minutes (1hour, etc..), will NOT work. (Hdds never go to sleep).

These results are for the Intel Rapid Storage driver 9.6 (iaStor.sys 9.6.0.1014) on Windows XP 32bit (SATA set as AHCI in BIOS) and Intel ICH9 controller.

Again, putting SATA back to IDE in the Bios, and there is no such bug: HDDs go to sleep regardless of the time set. (ie: 3 hours works OK)

-> iaStor.sys on XP will not allow HDDs to sleep if the time set for "Turn off hard disks" is beyond 45 minutes.

1. Could anyone with Win XP, AHCI enabled and Intel Rapid Storage driver 9.6 try to reproduce this bug?

2. How can I contact the Rapid Storage driver team at Intel to submit the bug? Are moderators able to forward a bug report?

Please let me know.

Here is the setting to put HDDs to sleep on Win XP:

[idata]

Intel? Anyone? ^^;

[idata]

Oohh..

It seems that "Intel Rapid Storage" version 10.0.0.1043 beta has solved this issue!I need to test more to make -sure- but HDDs did go to sleep after set time = 3 hours.. promising!

[idata]

Lucky you if this is your biggest problem with "latest" intel drivers.

My problem was (and still is I think) http://communities.intel.com/thread/15049?tstart=0.

PS: Do you know how to contact intel support?

[idata]

I have no idea how to contact Intel (please don't hijack this topic, I like talking to myself lol).

I sent a few private messages to the moderators (I think, the posters wih an Intel logo avatar?), but no answer so far.

Sorry I can't be of more help.

[DKnow2]

While it is possible to change IDE to AHCI post-OS-install using a command, I am not aware of any way to switch back to IDE safely.

[idata]

Actually there isn't anything safer than switching back to IDE (if you went from IDE to AHCI in the first place):

Just go in the BIOS, and set IDE. Heh. that's it.

Because Windows already installed the IDE drivers for that controller before, it will use them directly no questions asked.

After both drivers (for IDE and AHCI) have been installed at least once, switching from one to the other doesn't take anything more than going in the BIOS and switching there.

Edit: of course this is only valid for AHCI. Doing the same with RAID would be catastrophic!!

[DKnow2]

My concern is for Vista or Win 7, see

http://www.intel.com/support/chipsets/imsm/sb/CS-015988.htm http://www.intel.com/support/chipsets/imsm/sb/CS-015988.htm

http://support.microsoft.com/kb/922976 http://support.microsoft.com/kb/922976

I would expect XP to be worse but you are not experiencing the BSOD.

[idata]

"I would expect XP to be worse but you are not experiencing the BSOD."

I'm sorry but that doesn't make any sense. Of course XP will BSOD if you had it as IDE and suddenly switch to AHCI without preinstalling the driver; just like Vista and 7.

I'm not even sure what we're talking about here. In any case, this is MAJORLY off-topic, same for you wdc3003.

If you have no interest in the bug at hand, AHCI mode + iaStor.sys preventing HDD sleep, then please do NOT post on this thread, what's going on on these boards??

On a side note, I'm happy to report that IRS 10.0.0.1043 beta _DOES_ solve the problem

Because the timing in my reporting the bug versus 10.0.0.1043 release are very close, I believe an Intel IRS dev has seen this thread and managed to reproduce and fix the bug. To this person: THANK YOU!

So to sum up quick:

IRS version 9.6 had a bug on XP that prevented HDDs to go to sleep after the specified time in Windows' "Power options".

The new IRS version 10.0 solves this issue and lets HDDs go to sleep when they are meant to.

I hope IRS 10.0 WHQL isn't too far off!

[DKnow2]

I would feel more secure if you could vouch for the registry as correct, including any updates that might assume that either AHCI or IDE was enabled. I will check for myself next time. You only imply that AHCI is not enabled by default to allow for older drives and for faster startup. Please consider the diversity of viewers when posting. I haven't seen the BSOD and don't want to, much like most others in most forums. I'm glad your issue has been resolved but I have not yet seen hard drive sleep time as an issue for related Sleep/Hibernate problems.

[idata]

> I would feel more secure if you could vouch for the registry as correct, including any updates that might assume

> that either AHCI or IDE was enabled.

What on earth are you talking about?

In my case I went fom IDE to AHCI post OS installation. For this I had to install the AHCI driver in Windows before I could switch to AHCI in BIOS, otherwise Windows XP would bsod upon booting which is totally expected (can't find drivers for the controller in AHCI mode).Again, this is COMPLETELY OFF TOPIC.

> I will check for myself next time. You only imply that AHCI is not enabled by default to allow for older drives and

> for faster startup.

What? o_O

> Please consider the diversity of viewers when posting. I haven't seen the BSOD and don't want to, much like > most others in most forums.

Why did you even bring that unrelated BSOD issue in this thread?

> I'm glad your issue has been resolved but I have not yet seen hard drive sleep time as an issue for related

> Sleep/Hibernate problems.

Are you on XP? Are you using IRS 9.6? Did you even try to reproduce the bug this thread is supposed to talk about?

I'm afraid your answers were neither "Helpful" nor "Correct" so I won't flag them as such, which I now suspect is the sole purpose of your presence here.

[idata]

After a week+ of testing, I'm happy to report that Intel RST 10.0.0.1043 beta drivers definitely solve the "HDD don't go to sleep" bug on Windows XP 32bit.

Intel, thank you very much.

[BCumm]

help with Intel RST manager on I7-930/ASUS sabertooth, WIn server 2008R2, RST RAID 10 works fine, RST Manager will not load (" NETFrame 3.5 not loaded")/