VMware Communities
CuriosTiger
Enthusiast
Enthusiast

Sleep/wake issues - please report your experiences in this thread

There's been a lot of discussion about sleep/wake issues, but VMware is allegedly unable to reproduce the issues. I've offered them access to my logs and even to reproduce the bug for them on my equipment, which they have never taken me up on. I'm hoping that by consolidating everyone's reports in one thread, it may help get this issue the attention it needs.

To that end, my own experiences are below. I originally typed this up in response to a question in a different thread.

The Mac has two sleep modes:

  • Regular sleep, where the processor operates in minimal power mode but the RAM is still refreshed
  • Hibernation (or what Apple terms "deep sleep",) where the contents of RAM are written out to a sleep image and the computer then powers off

When you put a Mac to sleep, it initially enters regular sleep. Resuming from this state is no problem. But after a few hours (or when the battery drops below a certain percentage,) the computer hibernates. Waking up from hibernation is where the problem happens. The computer restarts, writes the contents from the sleep image back into memory, and you resume where you left off. But in some cases, Fusion interferes with this process, leading to a corrupted sleep image. And when the sleep image gets corrupted, the host OS will kernel panic on resume. That forces a full reboot, leading to the loss of all open work in all applications across the host and all guests. As far as bugs go, this is a pretty severe one. It has persisted across the last few versions of VMware Fusion (7.x and 8.x at least; not sure about 6.x.) Resetting the PRAM or the SMC has no effect. There are two known workarounds:

  • Using pmset to disable hibernation entirely DOES work, but with the drawback that the laptop will eventually run out of power while asleep without ever writing the contents of RAM to disk
  • Suspending all VMs before putting the computer to sleep also prevents the problem from occurring, even if VMware Fusion itself is still running and its various .kexts are still loaded

VMware has acknowledged this as an actual bug in the past, but no fix has been forthcoming.

6 Replies
dariusd
VMware Employee
VMware Employee

I feel your pain.

Here's a a slightly-reworked excerpt of my previous reply on this topic:

The short answer:  There is no change in [any recent version of Fusion] which is known to address the OS X sleep/wake issues which seem to correlate with the use of VMware Fusion.  That said, we still do not understand the root cause of the problem and still cannot reproduce the failure here, so it is not impossible (though I'd consider it unlikely) that we have somehow fixed it without realizing.

The longer answer is a story of a futile battle to troubleshoot a rare and complicated problem.  Please don't assume that we have been ignoring it... Our support stats show that a small but noteworthy proportion of our install base is affected, so we're very aware that there is a real problem here.  I can safely say that I've personally expended more effort on this one issue than on any other single problem I've ever worked on at VMware, and I'm just one of the team who've spent time on it.  We have an extraordinary team of engineers, but we're all pretty well stumped.  It doesn't help that the problem seems to be almost impossible to reproduce or to correlate with any specific configurations or actions, and that we really don't get any useful information out of OS X when it happens.  We had one Mac here at VMware which encountered a sleep/wake failure or blank screen on wake occasionally, right up until we started investigating it, at which point the problem went away – completely and permanently – despite our very best efforts to avoid making any change that could conceivably alter the problem's reproduction.  That was probably our best chance to make progress on understanding it...

I could continue telling the story of all the things that we've tried that have failed to lead us towards any sort of solution... but the sad conclusion is that we still have no idea what's going on, and no idea how (or if) it can be fixed.  We can't confidently say if it is a problem with VMware Fusion itself, or a problem in OS X that Fusion is somehow provoking, or even a problem with the hardware (or Mac firmware) itself, or something else entirely...

If you do happen to try [a recent version of Fusion], please feel free to let us know how it goes... but if you were previously affected by sleep/wake issues, I would not really expect an improved outcome.

I'll continue telling the story of all the things we've tried...

For a while, I had four Macs in my possession, and I was carrying two MacBook Pros to and from the office every day for no purpose other than to try to provoke a Sleep/Wake failure in any of them.  I tried all manner of different OS X versions, different guest OSes and workloads in the VMs, different "pmset" settings, and everything else I could think of to trigger the problem... AC versus battery, Wi-Fi versus wired network, frequent wakes versus 12-hour or 24-hour sleeps.  I spent an entire Memorial Day long weekend and a 4th of July long weekend, plus countless other weekends, with a total of four MacBook Pros, again doing little other than trying to provoke a Sleep/Wake failure on any of them.  I saw a total of one Sleep/Wake failure out of all of those Macs; It was running a debugging version of Fusion and a debugging version of Mac OS X, and even with all of the extra debugging turned on, there was zero additional information available... just the same single line in the host's system.log telling me that the "System was rebooted due to Sleep/Wake failure".

We've tried working with Apple to understand it, but since neither Apple nor VMware can reproduce the problem, and since there really isn't enough information to figure out anything further, that cooperative effort quickly ran out of opportunities to make progress.

I've produced a spreadsheet in which I've collated as much detail as I can find from customers who've filed Support Requests and customers who've reported this problem here in the forums.  There aren't any common factors that I can see.

I've made site visits in an attempt to find further any useful information from affected customers.

We've also had reports of the exact same crash when users are running VirtualBox or Parallels (and not running VMware Fusion) on the same systems.

  • Suspending all VMs before putting the computer to sleep also prevents the problem from occurring, even if VMware Fusion itself is still running and its various .kexts are still loaded

We've had a number of reports which contradict that claim, and state that the crash will occur even after exiting VMware Fusion.  This might be an accurate account of your own observations on your Mac(s), but please be aware that it is not universal.

VMware has acknowledged this as an actual bug in the past, but no fix has been forthcoming.

We've acknowledged that it is a problem.  If one of our representatives ever stated that it was a bug in VMware Fusion, that statement was made in error.  We are not currently aware of any issue in VMware Fusion itself which could be responsible for the Sleep/Wake failure.

We still have a problem report open to troubleshoot this issue and see that it gets fixed, wherever its cause may be and whatever the ultimate fix may be; I'm one of the engineers on the team assigned to resolve it.  By far the best way to make progress would be to get together, in one place: (1) one or more VMware engineers; (2) one or more Apple kernel/firmware engineers with deep knowledge of hibernation and wake; and (3) at least one Mac which reliably triggers the problem.

Cheers,

--

Darius

jbobbins
Contributor
Contributor

I have this issue a few times a week.  It's kind of getting tiresome.  I take my MacBook Pro home with me and don't usually immediately plug it in.  Those are the times when the problem typically occurs.  (I always report the issue with the "Report..." button.)  If I need to use it right away when I get home (usually not the case), I'll plug it in and I won't typically see the issue.

VMWare Fusion

Professional Version 8.0.2 (3164312)

Host:

OS X El Capitan

v. 10.11.13

MacBook Pro (13-inch, Mid 2012)

Guest:

WIndows 10 Pro v.1511

10586.839

64-bit

Let me know if you'd like any other information.

Eric_Allione
Enthusiast
Enthusiast

This happens to me on Fusion a few times in a row now and then, but it always works right eventually. This especially has been the case after Windows updates. I use Fusion every day for a Windows 10 VM. But it has been better behaved the past 6 weeks or so. Wher happens generally is a "red screen of death".

We seem to have slightly different causes and symptoms for this, but I would advise that you take the same precaution as me, which is to have 1-2 snapshots handy at all times. VMware recommends not keeping them older than 72 hours.

Another observation I get is that when I power off the Windows VM that it has a higher probability of issues. However, it almost always works for me if I hard kill the window by hitting the X and letting it take some kind of automatic hidden state snapshot. I wish I had a technical term for what happens when you kill a Fusion window like that, but it's much better on average than just leaving it and being at the mercy of Windows to install whatever it wants unattended. I also have just been shuttting down my MacBook most of time.

Reply
0 Kudos
enjayaitch
Contributor
Contributor

Me too.  Only on my iMac (Retina 5K, 27-inch, 2017, 4.2 i7, 32GB).  I've never had it on my MacBook Pro but I suspect I've never left Fusion running without closing VMs or suspending them.

No need to repeat the details - they are exactly as you describe - if I forget to close Fusion (or shut down or suspend the VM) last thing at night I can be sure the machine will have rebooted in the morning, with the sleep wake failure.

Since it's well over a year since your post this is all rather depressing.  Is there a formal channel through which we can at least register our experience directly with VMWare?

Thanks

Reply
0 Kudos
enjayaitch
Contributor
Contributor

Hello Darius

Your post is from quite a while ago - in case you're still at VMWare and still interested in this, I've pasted the reply I just got from your support team below.  The false syllogism is remarkable "I can't be a burglar because there was a burglary last week and it wasn't me."  Also suggesting the link (which was to this thread) says the same thing, which it doesn't.

So bitterly disappointing - the support response is infinitely more depressing than the issue itself; symptomatic of the descent into non-service that I see reported elsewhere on this forum.

Regards

Nick

Thank you for your Support Request.

As mentioned in the link as well that you have sent us.

This is an issue with Mac not with Fusion.

This issue was even found on those mac systems as well, where Fusion was not installed.

Still, if you have any queries, please let me know.

We will try our best to answer your questions with logical justifications.

Looking forward to hear from you.

Have a nice day.
Reply
0 Kudos
Gary777NJ
Contributor
Contributor

I've been a VMware Fusion user since version 7. I've experienced the dreaded "System was rebooted due to Sleep/Wake failure".

I upgraded to version 10 last year, now 11 again just this week. While I no longer get the above error under 10 or 11, I still have consistent issues. On a brand new MacBook Pro, both versions kill the internal display from waking coming out of "clamshell mode" (lid closed with MAC AC power in a supported docking station w/ external displays). I can recreate the issue EVERY time.

If I NEVER start VMware Fusion, the issue waking the internal MAC display NEVER occurs.

I'll try the option to suspend the VM first before the MAC sleeps the display to see if this is at least a workaround for me.

Reply
0 Kudos