I've seen on MacPerformanceGuide an issue with displays and sleep, and some MiniDisplay adapters - so I would say ATI 5770/5870, cables, display is one place to look.
And for that, probably SMC or/and EFI updates. And GPU drivers.
And anything dealing with powermanagement, reset SMC and delete any prefs.
If you have the time, might do a clean install (create a partition if needed) to test with.
Hard restarts can result in more trouble and corrupt disk directory.
I assume TTP 5.x was used and clean bill of health on drive and files?
Hi, I've done all the usual stuff, resetting the SMC and full diagnostics with TTP5 5.0.7 and everything is perfect. This particular machine is our media server but throughout the day is put to sleep many times. The failure probably occurs once every 50 sleep/wake cycles. If it were only put to sleep at the end of each day then the problem might occur once every 6 or 8 weeks.
On the hard reset issue I was assuming as it's asleep then the discs are synced on sleeping. I have certainly never had any corruption. I was having problems with an attached 1tb Icy Box firewire800 drive (time machine) which from time to time seemed to become unmounted. I thought that the problem could be something to do with that and so I reconnected the drive with USB. That has solved the unmounting problem but alas not the sleep issue.
I have a similar issue. My 2010 Exacore Mac Pro with 5870 correctly goes to sleep but sometimes doesn't wake.
I changed video card (using an "old" 8800 Gt) and the problem gone. AppleCare will send me a new card but I fear that the problem is not my card, but of all 5870 models.
Log file indicates:
Wake: Drivers Failure - AC - SATA IOU0
com.apple.message.signature: Drivers Failure
My guess: This is most likely just what it says... a driver problem.
Nonetheless, sometimes a particular setting or combination of settings can bring out problems that might otherwise go unnoticed during testing. So...
It would be helpful to folks here, at Apple, at ATI - If you posted your Sleep Settings. (ALL OF THEM, including Options) Maybe even your screensaver choice (if you have that set to go on sleep)
Also, Try changing settings 1 at a time (go back to original then change another) for example: disable HD sleep.(Personally, I NEVER use hard drive sleep on non-portables anyway), disable "Wake for Ethernet" If something works Report back AFTER a decent period of time.
Obviously changing those settings, won't fix the problem but MIGHT allow for a workaround until Apple can fix.
One last temporary preventative: Set Put the computer or monitor to NEVER sleep Shut your computer down when your not using for more than a couple hours.
Lastly, notes on when this began occurring (ie after a particular install) can occasionally be helpful (particularly if it's NOT a driver problem)
Message was edited by: Alix Leigh
Message was edited by: Alix Leigh
Do you have any updates on this issue?
I have the exact same issue on my 2012 12-core Mac Pro with ATI Radeon HD 5870. I have the stock 12 GB memory, I'm running four 1TB HDs in software RAID 10 (1 + 0), with four 27" Apple LED displays. I am running Mac OS X: 10.6.8
One time in ten, or so, I'll put the Mac to sleep, the fans will spin up, the light will flash like it's sleeping, but the fans will never spin down, and it's impossible to wake the Mac. I am forced to reboot at this point.
In my Power Management logs (/usr/bin/pmset -g log) I see the following sort of errors (the only thing that changes is the date):
* Domain: sleep
- Message: Sleep: Drivers Failure panic - AC - IOU0 HDEF SATA
- Time: 9/5/12 11:05:48 PM MDT
- Signature: Drivers Failure panic
- UUID: 2391C1EB-1D5B-4530-8FA9-42D2430C3D16
- Result: Failure
I'm curious if this is the ATI card as I have historically had bad experiences with ATI and Macs (as well as the above comments from Carolus730), and video cards do have a higher than normal instance of causing kernal panic issues...
Sleep: Drivers Failure - AC - UHC6
UHC6 is one of your USB connections (unless another driver happens to use that name, but it would be bad assumption). There are applications to show the I/O tree (which includes the USB branch), particularly if you install Developers Tools, so you have some clues as to which device is the culprit (sometimes, a clear name appears).