After installing mavericks, the iMac will not wake from sleep properly (at all)

Apple has nothing to do with these discussion boards. The only way to plug away at Apple about the problem is to use the feedback page.

After doing what all of you have probably already done (exhasitive trial and error), the only thing I have found that works (other than shutting off the computer at night) was an equally ridiculous solution.

Energy Saver: Never and Never.

Screen saver to blank screen after 10

The past two mornings are the first two in a long time when my computer wasn't locked in sleep mode.

I know, I know....but its something.

Nixclevername, actually not all that bad. On a laptop with battery power sleeping the processor is important. But on a desktop like the iMac which runs on AC not being able to make the CPU sleep is not a big problem.

I do think though that I want my displays to be able to sleep. In my case the 27 inch iMac screen and two 24 inch LCD monitors. They generate a fair amount of heat (reflecting energy consumption), but mostly I want to preserve them, working well as long as I can. Being powered down with the screen illumination off and no heat load contributes to that.

To that end I was pleasantly surprised to find the Shift-Control-Eject keyboard function actually does that, rather than trying to sleep the whole machine, which some of the sparse documentation had suggested to me would be the case. The displays go dark immediately and power down (I can tell because both external monitors have lights that show they are powered down, not just displaying black screens with the illumination on behind them). The displays also come back nearly instantaneously when I hit a key. Both the shutting down and coming on are much faster than what I see when the machine does a full sleep and a full wake from it.

David Neale1, in general I agree, but there are some things to keep in mind.

These discussions are on a site run by Apple, not some random place on the Internet. So it is not unreasonable to think they might serve as a resource for Apple maintainers if their attent is drawn to them - say through the feedback path and maybe through AppleCare cases for example.

I think maybe there is another path that alerts Apple to problems that has not been mentioned. I have been monitoring my system's logs using the Console utility. I see a good bit of activity logged by crash reporting. In particular it seems there are crash reports being generated and sent off by WindowServer that seem to be related to the apparent failure to wake symptoms. I am not quite sure where you control it after the initial question, but I suspect the crash reporting is at least some of what we allow when we are asked if we want to share information (anonymously) with Apple and we say yes.

Judging from my logs and crash reports WindowServer seems to be at the heart of the functions where things are going wrong, at least in the iMac/Mavericks situation, at least in my situation.

I have done a good bit of development and maintenance over the years and I have to keep in mind the challanges Apple has. There are millions of Macs out there and every one has different combinations of hardware, software and usage. And most of the users are not sophisticated and knowladble professionals who can provide the level of input the maintainers need in a case like this that appears to involve complex, unanticipated interactions.

There is one more point too. Watching the various people post in this topic I see what seem to be conflicting observations that make zeroing in on answers difficult. Many users cite a problem that came with Maverics, but there a few saying it is in 10.8 as one example.

I think the reason for the diversity and conflicting observations is the way this forum works and is used. You come here and search on something like "will not wake" and end up in a topic like this. Different underlying problems end up converging into a single topic because they share a symptom.

For example not waking seems to be a symptom people observe in many different situations. I just noticed another topic that started a few years ago, basically about not waking, but obviously from long before Mavericks. There must be a variety of causes at different points in time and different sets of circumstances that end up with the observed symptom of will not wake.

So, I think it is helpful to keep in mind we may not all be looking at the same problem and that is one of the reasons we see different things.

My sense is this problem - will not wake, arriving with Mavericks - is a complex problem. There were a number of changes in 10.9 and somewhere there are incompatibilities with some hardware or software or usage patterns that interact to cause some of us to see a "will not wake" symptom. I am guessing that it has a timing component - "race" conditions for example - during the time the system is storing machine state and shutting down and/or when restoring state on wake. In my particular case I either have more than one problem, or the underlying issue can manifest in different ways. Most typically after sleep (typically over night, but not always) hitting a key just does not wake - screens stay dark and nothing noticible happens. But sometimes I see a gray screen come up. Some times the external monitors seem to have not been powered down because I see indications they are powering down when I am trying to wake. Most of the time the logs show the machine continued a level of activity (as it needs to) during sleep and was trying to restore state and wake up until I finally did a hard shut down, but in at least a few cases the logs seem to show the machine stopped when it went to sleep - it stopped loging completely. I also see what seem to be a lot of log entries around the times of these issues that suggest inconsistencies in what has been restored - window ids wrong, can't find something, etc. etc.

All that suggests to me some sort of process that is causing memory corruption. My guess right now is that some of the changes in screen handling, or app nap, or the aggregating of background operations of different apps or some such has tripped up assumptions in some code about how things work and their timing, or exposed latent errors that were there before but did not cause a problem in the past, and only surface in certain combinations and conditions under 10.9.

Excellent note Bob! Well written note on the usefulness and role of discussion groups. I'm going to bookmark this as a reminder in other forums whenever expectations get a little whacky (not that that's the case here at all). Thank you.

I had a similar issue with my MacBook Pro (mid-2012) running Mavericks on my Thunderbolt Display (I know this is about iMac issues, but bear with me a little and the troubleshooting steps may be useful in any of your iMac-specific circumstance).

The symptoms:

• the MacBook Pro was asleep, attached to the Thunderbolt Display and when any key is pressed it wakes up as expected.
• the MacBook Pro was asleep without attachment to Thunderbolt Display, any key would wake it up normally, as expected.
• the MacBook Pro was asleep, attached to the Thunderbolt Display, detached from the Thunderbolt Display while still asleep, and then wake attempted - it would not wake up normally. Instead, the MacBook Pro would show activity LED when detached and a few moments later, without opening up the display, when I was walking out of the house with the MacBook Pro in my knapsack, it would do the POST chime. Not an expected behavior.

The steps I took:

• tested with known good OS. Built a Mavericks startup disk on a spare external hard drive. Known good OS did not exhibit the third symptom. This performed as expected and showed that the problem was with my software, since both the known good OS disk and suspect hard drive shared the same hardware.
• tested with another user account. Problem existed. This told me that the issue was with system-level software.
• wiped and reinstalled the OS, restoring from Time Machine. (I wish I hadn't done this one quite yet, because it merely delayed the inevitable item-by-item search of the next step). Problem persisted.
• read this discussions.apple.com article and downloaded the EtreCheck utility software. Ran it on each of the suspect OS and known good OS to and then took steps to remove each of the reported elements that were different. After removing one item, I would test again. If problem persisted, I would remove the next item on the list and so on. Even if I thought it was not likely a particularly popular plugin, I took it out and tested it to be sure I was systematic. Besides, I could always load the plugin back.
• I struck gold with the kernel extensions, and in particular a VPN-related extension from client software I had tested six years ago. This was the culprit.

Now, the MacBook Pro runs like a champ without odd sleep-wake issues.

My two ¢:

1. build a known good OS and hold down option to select it when restarting the iMac (This shows all the possible boot drives). If the problem persists with the known good OS, then it's likely a hardware issue. Otherwise, it's a software issue with either your account or the OS.
2. Test in another user's account. If problem persists here, it's likely the OS.
3. You can wipe, reinstall the OS and then restore from backup, but this can be wasted effort like it was in my case since the issue persisted.
4. Compare the EtreCheck report from the known good OS versus the existing OS. Use this list and Google searches to discover the method to remove plugins or extensions or whatever you're trying to take out of the iMac. Trust me, you're not the first Mac user to ever try to remove/uninstall something specific. Chances are that not only are you not the first, but enough have had issues that they wrote about it fora like this one.

I think I finally have something useful to report. My wife's iMac 27" Early 2011 started exhibiting the same symptom after 10.9.1. She has attached to the iMac a combined USB2.0/FW400 "Moshi" brand hub. I discovered that by unplugging her external firewire drive cured the sleep problem. I also noticed that the drive was failing. So I replaced the FW enclosure with a USB 2.0 enclosure and plugged it into the Moshi Hub. After a couple of hours, the black screen re-appeared. I then used disk utility to repair permissions and repair the disk (it has 2 x 500Gb partitions) and the blackness was still apparent. I then took the USB lead out of the Moshi Hub and plugged it directly to the iMac after taking her scanner lead out and putting it in the hub. I then restarted and performed an fsck (the volume was modified) .So far, the problem has disappeared. I'm not going to comment further because a few days ago I made some pretty abrupt suggestions to our patron and the whole item was smartly removed. I was encouraged to send feedback via the proper channel which I did so hopefully someones inbox is piled high with requests for a fix.

I have not been actively following this thread as much as others here on the forum that seem to describe the exact same problem, but I think I agree with the implication of your finding. Not everyone describes all the extra components they have connected to their systems, but in this and the other two threads, it appears that many have either an extra monitor or a USB connected drive. Though I'm getting (or was getting before turning off power saving features) the problem on a Mac Pro that has both extra devices plus a firewire soundcard, the mini server I had that exhibited the problem, had only a USB connected external drive. Perhaps this problem relates to non-apple hardware being connected when the computer attempts to come out of hibernation. That would be less likely for them to test with and that WD USB drive issue was a major problem in the initial installations of Mavericks, there is possibly a precedence and lesson set there about testing of "other" devices. It doesn't seem like a stretch that Mavericks may be having wake-up issues with "unexpected" devices that were not tested.

Another reflection. perhaps a bit fluffy and OT... If the mods are deleting threads as you observed, hopefully they are reading these suspiciously similar threads and telling someone who can recreate the symptoms. Moderators, are you listening 😉?

In response to: Option + Shift + Eject.

Worked like a charm. So. My settings are NEVER, NEVER with screen saver at 10 mins (blank) and then at night, I leave and PUT SCREENS TO SLEEP. 🙂

Just a quick note on one of my previous posts:

I did recover a late 2012 Time Machine random backup and the issue had no problem, no sleep. (iMac Mid 2011, 16 GB). I've not tried to recover post Maverick but pre 10.9.1 and I have not tested pre Maverick upgrade. (lazy at this stage, so many tests, trial and error.). Just fyi...worked with old recovery.

I think my most extreme test causes me to assume it is not hardware or external items.

Format machine.

Rebuild from ground zero

Load Mavericks.

As close to factory as possible.

Not a single item plugged in or app loaded. (iMac Mid 2011).

Apple ID login only for purpose of RESTORE internet. No options selected.

Ground ZERO was only option I could do.

Answer: Overnight, bright gray screen, no recovery. Bluetooth keyboard no work. Then tried same event after a couple nights but added a plugged keyboard. Same results.

Side note: I have 3 iMacs sitting in front of me (yes, and don't ask). I have a MacBook Pro on desk behind me. (again...don't ask. Just assume I'm crazy about apple). At home I have 4 operating "family" iMacs from late 2010 to current. All have Maverick. Two are 27" and one is the smaller ??". Ready for this? NO ISSUE. I can't replicate the problem on any machine other than THIS iMAC. Mid 2011. Maybe I'll post my system info and someone can compare. Maybe there is a commonality between a very specific hardware config that we are missing.

For purposes of OTHERS comparing. The most confusing part of this to me is my inablity to replicate the issue in my OTHER iMacs. I have 3 at work and 4 at home. All ranging from late 2010 to current. This iMac is the ONLY one with the "Maverick" issue which is confusing.

iMac 27 inch, Mid 2011

Processor: 3.4 GHz Intel Core i7

Memory: 16 GB 1333 MHz DDR3

Graphics: AMD Radeon HD 6970M 2048 MB

Software: OS X 10.9.1 (13B42)

Side note: In the console I can't make any connection at the time of last action besides it always happening around 3:30 am to 4:00 am. Tries to come alive around two hours later. Bunch of "KERNEL" stuff that I don't understand. Doubt this helps. I am just hoping there was a random set of machines loaded with a specific combination of system that is in conflict. Oh well...just a thought.

Yes, replication is a real problem, even on the same machine!

For the past few days, the iMac 27" Mid 2010 with which I have had the problem has been reasonably well behaved. The problem was never 100% consistent (sometimes a black screen, sometimes a white one, sometimes a good screen but no further response…) and recently a new strangeness has appeared: the iMac wakes up, but does so with at least one "transparent" window (usually Safari), where only the edges are visible, but with some other windows correctly overlapping and others visble through the Safari window. Very strange. At times, too, an area of graphic "interference" appears in a section of an apparently random window when waking, but it can be removed by scrolling that part of the window out and back again, or by first overlapping with another window.

I'm sure all this has to do with the same sleep/wake issues: as BobHassinger proposed above, this is not a single issue, but a number of different ones that manifest themselves with the same symptoms.

Hello folks,

I'm affected with this problem. In my case, I have a 27" iMac Late 2009.

System is not allowed to go to sleep, only the displey is confirgured to sleep.

Symptoms are when the display goes to sleep, time to time I can't turn it back on.

Yesterday, I think I have found a way to reproduce the problem, consistantly, if the system is heavy loaded (like a video file conversion) for a while, I consistantly can't wake the display up.

Though the iMac can be access from the network (ping, SSH), so the OS is still running and the problem is the graphic interface process or the video driver.

Quesiton for those who experience some improvements with the nvram/other reset, disc permissions repair and all.

Did you have any luck with this ? Did it fix the problem ?

Thanks,

Johan

Those did not help me, and I don't recall many reports from others where it stayed fixed, even if they thought it had helped initially.

Your observation Johan is pretty much consistant with others, including mine where you will get a wake after a short sleep, but particularly over night the problem seems to show up more often.

I continue to watch the logs with Console and try to tease out clues. There is a task/process/whatever OS X calls it named WindowServer that seems to pretty consistently report crashing around the problem, and I tend to see a lot of messages about invalid display IDs and the like. I suspect that as copies of the current displays and windows are captured around going into sleep, and used to restore the information for them on wake, that some part of the information is being lost or corrupted.

At the moment I am keeping an eye on iTunes and the various ways I use it. There are often log entries from it around the time when the displays are getting into sleep/wake trouble. I use it to play podcasts on a full screen mostly, and I usually have my iPhone plugged in over night so there may be intreraction with it, or its coming and going. My machine has three screens and the new handling of screens introduced in Mavericks is exercised in that. I noticed that there was an iTunes update a few days after Mavericks came out that was supposed to address Mavericks compatibility. It is not totally out of the question that it brought a problem with it, rather than the issue being with Mavericks itself I guess.

I don't r ecall seeing the problem for a few days after the Mavericks install so I am wondering if anyone saw this problem for the first time after iunstalling Mavericks (10/25/13 for me) and before the iTunes 11.1.2 update (11/6/13)?

Next time the display would lock up, I would watch the status of the WindowServer process.

Just food for though, would it be possible to restart the process if it is crashed ?

Here is the full path : /System/Library/Frameworks/ApplicationServices.framework/Frameworks/CoreGraphic s.framework/Resources/WindowServer -daemon

If the process is still in memory, I would definitly try this : sudo killall -HUP WindowServer

Here is a peek in my logs :

imac:log root# cat system.log | grep -i windowserver | more

Jan 27 00:25:19 imac.yhan.ath.cx WindowServer[87]: _CGXHWCaptureWindowList: No capable active display found.

Jan 27 00:56:14 imac.yhan.ath.cx WindowServer[87]: _CGXHWCaptureWindowList: No capable active display found.

Jan 27 01:21:04 imac.yhan.ath.cx WindowServer[87]: _CGXHWCaptureWindowList: No capable active display found.