Q: Dual Quad-core Mac Pro reboots from sleep

+... take two aspirins and call us in the morning.+

Seems like you have to shutdown every night, and unplug and leave off line, effectively draining any trickle current, and doing an SMC reset, daily. ;-(

+... but under no condition say, "we don't know why."+

shameful, Apple's muteness is shameful...

I am having this problem too. (8 core 2.8, stock 2GB, 8800GT). I have also noticed a few other issues...

- The system does not seem to like to boot with my drobo attached.
- After doing the graphics update the other day it hung on the grey screen after the bong. I probably left it for a good while (I went to make a cuppa) and I thought I would remove all my usb devices and force a power cycle - as soon as I pulled my iPod shuffle USB cable out the machine continued to boot.

- I have also had the keyboard and mouse not work after a boot up. First time I unplugged my monitor USB (Dell 24" with built in card reader, hub etc) and reconnected and they started working. Second time I just left it for a really long time and then it worked!

My gut says some sort of USB problem!

Reboot but when you hear the boot beep, hold the keys cmd-v down until you start to see text on the screen. You'll be booting "verbose" so you can see all of the messages in the system. When it gets stuck you can see what it's doing or the last thing it did.
You may be hanging on a drobo scan but this should give you some indication...

HerrDrKarl wrote:

Idle with 2 x 2 GB RAM installed: 195-199 W
Idle with Apple 2 x 1 GB RAM only installed: 175-180 W

Idle with 4 x 2 GB OWC RAM installed: 212 W

I have had zero problems since beginning these memory experiments. With the current 8 GB (4 x 2 in slots 1 and 2 of each riser) I ran Apple Hardware Test overnight looping in deep mode (cmd-L to turn on looping before clicking the Test button). No problems found. Ran rember from desktop while I was busy this afternoon, yes, would have been better to do memtest from the command line in single user mode. But, no errors.

I'm about to be brave and go back to my 'problem' configuration which was with all 10 GB in place.

I did notice from updated comments on barefeats.com that I did not have my memory installed 'optimally' before - when I was having all of the reboot problems, GPU, and crashing.

Silly me, I followed Apple's directions (and those in the OWC box) which said:
1a: move one DIMM from bottom card to top card - with an illustration showing the Apple stock DIMMS in slots 1 and 2 of the top card
1b: put the first pair of additional DIMMS in slots 1 and 2 of the bottom card
2: but the second pair of additional DIMMS in slots 3 and 4 of the top card

So, I had Apple 1 GB DIMMS in slots 1 and 2 top and OWC 2 GB DIMMS in slots 3 and 4 top and slots 1 and 2 bottom. It seemed odd, but I followed the instructions anyway.

Barefeats.com updated their memory article on Feb 22:
http://www.barefeats.com/harper3.html
and it now has a note from an engineer stating "The pairs on the top riser (slots 1&2 or slots 3&4) should match the pairs in the bottom riser (slots 1&2 or slots 3&4) in terms of size if you want the maximum benefit to be gained by interleaving in real world applications."
(Please follow the URL to the actual article as they will no doubt be continually updating it.)
Well, this seems to make sense - and says that I installed my memory incorrectly originally. If true, it makes you wonder why the Mac Pro user manual says nothing about this - and that there is no diagram / info on the Apple support site, as there was for the earlier versions of the Mac Pro.

So, I'm about to put the stock Apple DIMMs back in, but in slots 3 and 4 on top this time. I'd like to think this is somehow the problem, other than that 'sailhome' has exactly this configuration of OWC and Apple RAM and is still having problems:
http://discussions.apple.com/message.jspa?messageID=6711026#6711026

Otto's experience with Aperture is also interesting to read ... and makes you think this is all a software issue. But, then I have to wonder why the problem disappeared for me with different memory configurations but with the same daily software load. Very odd situation we're all in!

Will post if anything interesting happens...

Karl

Wow! Perhaps this would explain why I could not replicate the problem with the stock Apple RAM installed, but began to have it immediately after I re-installed my TransIntnl RAM.

Like yourself, I simply followed the instructions in Apple's user manual. I will try what the barefeats article suggests and install my Trans. RAM in 1&2 on the top and bottom, and Apple's RAM in 3&4 on the top. Hopefully this will help - I'll report back either way.

I too have been bit by this problem on my brand-new Mac Pro with single quad-core CPU. It has a 2600 XT, 6 GB of RAM (4 add'l from OWC) and four internal SATA drives. I'm going to change the hibernation mode to see if that makes any difference. The default value of hibernatemode on desktops is 0 which means it won't save memory to disk. I used this command to change it:

sudo pmset -a hibernatemode 1

So now I have a 6GB sleep image in /var/vm/sleepimage. I don't really recommend other folks try this. I'm just curious as to whether storing the memory state to disk ("safe sleep") rather than keeping it in RAM might avoid whatever strange issue causes this problem. I'll report back in a few days on what effect, if any, this has.

Here is an extract from the pmset man page which talks about hibernatemode:

SAFE SLEEP ARGUMENTS
hibernatemode takes a bitfield argument defining SafeSleep behavior.
Passing 0 disables SafeSleep altogether, forcing the computer into a
regular sleep.

0001 (bit 0) enables hibernation; causes OS X to write memory state to
hibernation image at sleep time. On wake (without bit 1 set) OS X will
resume from the hibernation image. Bit 0 set (without bit 1 set)
causes OS X to write memory state and immediately hibernate at sleep
time.

0010 (bit 1), in conjunction with bit 0, causes OS X to maintain sys-
tem state in memory and leave system power on until battery level
drops below a near empty threshold (This enables quicker wakeup from
memory while battery power is available). Upon nearly emptying the
battery, OS X shuts off all system power and hibernates; on wake the
system will resume from hibernation image, not from memory.

hibernatemode is set to 3 (binary 0011) by default on supported porta-
bles.

hibernatemode is set to 0 (binary 0000) by default on supported desk-
tops.

Please note that hibernatefile may only point to a file located on the
root volume.

The adventure continues, but [over here|http://discussions.apple.com/thread.jspa?messageID=6721146#6721146] so as not to hijack this thread.

-sailhome

sudo pmset -a hibernatemode 1

OK, so this failed completely. On the very next wake I had a reboot

Yeah, but...
we already know that even if you have ALL banks populated with all OWC, or ALL Apple memory, the issue occurs, as well as with stock apple ram occupying only some slots (but in the correct interleave/order).....

Well, I reconfigured my memory as the BareFeats article and HerrDrKarl suggested, and my MP booted from sleep the very first time I put it to sleep!

Another potential "solution" fails.

Update: after reconfiguring my memory as suggested by the Bare Feats article, I've begun to have systemwide problems on my MP: spinning beach balls, apps unexpectedly quitting, and a variety of other strange issues.

This is disconcerting since, other than the reboot from sleep problem, my Mac Pro ran perfectly before I switched the RAM configuration. Is it possible that this change in configuration could be causing these problems?

Message was edited by: sw_i_t_t_e_rs

What reconfiguration did they suggest, can't find where this article is.

sw_i_t_t_e_rs wrote:
Update: after reconfiguring my memory as suggested by the Bare Feats article, I've begun to have systemwide problems on my MP: spinning beach balls, apps unexpectedly quitting, and a variety of other strange issues.

This is disconcerting since, other than the reboot from sleep problem, my Mac Pro ran perfectly before I switched the RAM configuration. Is it possible that this change in configuration could be causing these problems?

Message was edited by: sw_i_t_t_e_rs

Switter - I found where you posted the suggested configuration. Thanks

You possible damaged a stick. Try them a pair at a time. Run a RAM check after each reconfiguration... tedious I know.