Windows 10 VM pauses periodically, ignores mouse/keyboard

Discussion in 'Windows Virtual Machine' started by RodB, Feb 27, 2017.

  1. Dmitry@Parallels

    Dmitry@Parallels Parallels Team

    Messages:
    702
    @KorkyPlunger, I will send you instructions in a personal message with the instructions to proceed. This situation is pretty unique.
     
  2. RodB

    RodB Member

    Messages:
    62
    @Dmitry@Parallels : I'd not been working on this system for several days. Today I am again on it and just had an event for which I sent technical data #173684393. Time was just before 10:00 am.

    I was working in SQL Server Management Studio and it became unresponsive then, as usual, after a few seconds, again began working.
     
  3. Dmitry@Parallels

    Dmitry@Parallels Parallels Team

    Messages:
    702
    Hi @RodB, I've looked into the new problem report, and again there is no errors and vCPU stats are good:

    >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Timings <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

    ==== VCPU 0 ====

    Host = ( 0%) 12634953
    Monitor = ( 1%) 15856849
    Host HLT = (85%) 1228445115
    Monitor HLT = ( 0%) 2294988
    Native guest = (12%) 181891145
    Accumulation counters:

    ==== VCPU 1 ====

    Host = ( 0%) 9305124
    Monitor = ( 0%) 11827030
    Host HLT = (87%) 1298649356
    Monitor HLT = ( 0%) 2261459
    Native guest = (10%) 157143796
    Accumulation counters:

    ==== VCPU 2 ====

    Host = ( 0%) 4961244
    Monitor = ( 0%) 8046683
    Host HLT = (95%) 1419120527
    Monitor HLT = ( 0%) 2058160
    Native guest = ( 3%) 47097544
    Accumulation counters:

    ==== VCPU 3 ====

    Host = ( 0%) 2854950
    Monitor = ( 0%) 8942600
    Host HLT = (95%) 1416536484
    Monitor HLT = ( 0%) 2816152
    Native guest = ( 3%) 46509862

    -----
    Approximately 95% of all computer resources are free, so we're dealing not with the performance issue. The only suspicious messages that I've found are in system.log:

    Apr 19 09:59:44 por-barnesr-m3 McAfee: [101]: Alert: LogTime: 2017-Apr-19 09:59:44 AVAS: PID: 42631 : curl_easy_perform() failed: Timeout was reached, CurlErrorNo: 28,HttpErrCode: 0
    Apr 19 09:59:44 por-barnesr-m3 McAfee: [101]: Alert: LogTime: 2017-Apr-19 09:59:44 AVAS: PID: 42631 : GTI Query Failed!!
    Apr 19 09:59:45 por-barnesr-m3 McAfee: [101]: Alert: LogTime: 2017-Apr-19 09:59:45 AVAS: PID: 42633 : curl_easy_perform() failed: Timeout was reached, CurlErrorNo: 28,HttpErrCode: 0
    Apr 19 09:59:45 por-barnesr-m3 McAfee: [101]: Alert: LogTime: 2017-Apr-19 09:59:45 AVAS: PID: 42633 : GTI Query Failed!!
    Apr 19 09:59:49 por-barnesr-m3 McAfee: [101]: Alert: LogTime: 2017-Apr-19 09:59:49 AVAS: PID: 42631 : curl_easy_perform() failed: Timeout was reached, CurlErrorNo: 28,HttpErrCode: 0
    Apr 19 09:59:49 por-barnesr-m3 McAfee: [101]: Alert: LogTime: 2017-Apr-19 09:59:49 AVAS: PID: 42631 : GTI Query Failed!!

    I suggest the following:
    1. Disable McAfee on Mac and check if it changes anything.
    2. Look closely inside Windows VM: check event viewer after the freeze, check for available updates.

    By the way, similar issue happens on real PC. Maybe it's an issue with Visual Studio in the first place:
    https://developercommunity.visualst...s-2017-stops-responding-freezes-locks-up.html
    http://stackoverflow.com/questions/4607640/visual-studio-2010-periodically-hangs-for-several-seconds
     
  4. RodB

    RodB Member

    Messages:
    62
    @Dmitry@Parallels - Thank you. Interesting links you shared about VS. I've been suspicious about VS and SSMS (which relies upon essentially the same IDE) from the the beginning. However, on 4/11, I was testing this and had it happen when running only Notepad++ without VS or SSMS open.

    Well, I guess it is just something we'll have to live with as annoying as it is. I have spent hours and days trying to nail down when it occurs and sending a report but that has been fruitless. It is just too difficult to catch it at the moment it is happening. By the time I get the report generated, the event has ended, so I never know if it even captures anything useful.
     
  5. RodB

    RodB Member

    Messages:
    62
    @Dmitry@Parallels - Well, let's try this one more time. I have nothing open in the VM except for the vSphere Client. Neither VS nor SSMS is open, nor have I used them at all today since I booted the VM. I just had an event occur about 10:55 am and generated a report just after the event ended. Report #178823830 sent to your attention.
     
    Last edited: May 8, 2017
  6. RodB

    RodB Member

    Messages:
    62
    @Dmitry@Parallels: I just had the same event with the same conditions. Sent report #179079358. (Unfortunately, I happened to hit [Enter] in another app just as the dialog finished and it captured my [Enter] to send the report without any description.) I had only vSphere Client running. No VS or SSMS was running. Nothing had been used on the VM since boot was the vSphere Client.

    One very odd thing: The time was the same -- about 10:50 am. That seems an unlikely coincidence.
     
  7. RodB

    RodB Member

    Messages:
    62
    @Dmitry@Parallels: Just had another one with the same conditions. Time was about 1:25pm. Report #179105843
     
  8. Dmitry@Parallels

    Dmitry@Parallels Parallels Team

    Messages:
    702
    Hi @RodB, thank you very much for your reports. I have looked through our logs but couldn't find any errors, though I have a hunch:

    1. Mac logs are flooded with the messages:
    May 9 10:47:59 por-barnesr-m3 com.apple.xpc.launchd[1] (com.apple.quicklook[82100]): Endpoint has been activated through legacy launch(3) APIs. Please switch to XPC or bootstrap_check_in(): com.apple.quicklook
    May 9 10:48:27 por-barnesr-m3 xMetering[126]: Number of registered processes: 330
    May 9 10:49:27 por-barnesr-m3 xMetering[126]: Number of registered processes: 329
    May 9 10:50:27 por-barnesr-m3 xMetering[126]: Number of registered processes: 328
    May 9 10:51:27 por-barnesr-m3 xMetering[126]: Number of registered processes: 329
    May 9 10:52:27 por-barnesr-m3 xMetering[126]: Number of registered processes: 328
    May 9 10:53:28 por-barnesr-m3 xMetering[126]: Number of registered processes: 329
    May 9 10:53:53 por-barnesr-m3 com.apple.xpc.launchd[1] (com.apple.CoreRAID[82683]): Endpoint has been activated through legacy launch(3) APIs. Please switch to XPC or bootstrap_check_in(): com.apple.CoreRAID
    May 9 10:54:04 por-barnesr-m3 com.apple.xpc.launchd[1] (com.apple.ReportCrash.Root[82742]): Endpoint has been activated through legacy launch(3) APIs. Please switch to XPC or bootstrap_check_in(): com.apple.ReportCrash.DirectoryService
    May 9 10:54:28 por-barnesr-m3 xMetering[126]: Number of registered processes: 336
    May 9 10:55:28 por-barnesr-m3 xMetering[126]: Number of registered processes: 337
    May 9 10:56:28 por-barnesr-m3 xMetering[126]: Number of registered processes: 333

    2. I believe that xMetering belongs to FireAMP. Let's try to turn it off and check the issue.
    3. Also, to localize the issue, it would be useful to try another VM on the same Mac, to make sure that this issue is not VM-related. If it's possible, please install a new VM (you can download Windows Test Environment via PD Wizard: http://kb.parallels.com/en/122641), test the issue and let me know the results.
     
  9. PragmaticPraxis

    PragmaticPraxis Bit poster

    Messages:
    74
    Rob,
    I am seeing the same freezing issue, and in my case I am not using, or even installing, Visual Studio. I am using PowerShell ISE, and NotePad ++. That last commonality has me intrigued. FWIW, I have seen in the issue on multiple VMs, both Win7 & Win10, and the Win10 VMs are not on Creators Update.

    Dmitry,
    I tried to send a problem report, but I get an error about network settings. And yet, here I am on the Forum so obviously not a network issue. I tried attaching the report here, but the web site won't accept a file that big. Any suggestions on how to get this report to you?

    If there is anything else I can try to expand the data, or even verify Rob and I are seeing the same problem, let me know.

    Gordon
     
    Last edited: May 10, 2017
  10. RodB

    RodB Member

    Messages:
    62
    I have a Win7 VM I will use today instead of the Win10 VM. I have never seen the pause on the Win7 VM. However, neither have I used it as much since I got the Win10 VM. Plus, FireAMP was installed only recently on my system -- IT is doing a beta of it. I'll first try the Win7 VM leaving FireAMP in place and see if I experience any pauses. Depending upon that outcome, I can look into getting my system out of the FireAMP test.
     
  11. PragmaticPraxis

    PragmaticPraxis Bit poster

    Messages:
    74
    I just had another freeze. Windows 7, and no dev tools running at all, just installing Autodesk Revit 2015. Problem report 179601257 uploaded ok though.
    Gordon
     
  12. PragmaticPraxis

    PragmaticPraxis Bit poster

    Messages:
    74
    And another freeze. Restarting the VM doesn't seem to help this time. Uploaded a problem report with a dropbox link to a DMP file. How many more times do I need to do this before Parallels shows some RESULTS. This is getting seriously ^#$# frustrating.
     
  13. PragmaticPraxis

    PragmaticPraxis Bit poster

    Messages:
    74
    And AGAIN. When does this actually start to embarrass Parallels? And to be clear, disk optimization is NOT enabled. On any of my VMs, even the SSD hosted VMs where it in theory could/should be.
     
  14. RodB

    RodB Member

    Messages:
    62
    @Dmitry@Parallels - Just had another event and generated report #180798973 immediately after. Time was approximately 9:39am local. We've made some changes on my system regarding FireAMP and the Snow Inventory software in an attempt to better isolate this issue. Please review the logs and see if you notice any change.

    NOTE: I never experienced a pause on my Win7 VM but I did not have an opportunity to use it very much. A thought, however: I've had my Win7 VM for a very long time, at least a couple years. The Win10 VM is relative new, only a few months old. Don't know that has anything to do with anything but just trying to isolate considerations.
     
  15. RodB

    RodB Member

    Messages:
    62
    It certainly seems like this even occurs more frequently when using Visual Studio on the Win10 VM. But, it is possible I just notice it more because when I'm using Visual Studio, I am typically working solely on the VM as opposed to other times. However, there have been multiple times when Visual Studio is not open nor even been open since boot and the event occurs using other applications.
     
  16. RodB

    RodB Member

    Messages:
    62
    @Dmitry@Parallels - How do we move forward to correct this? It is not getting better, only getting worse. It is becoming increasingly frustrating when I am trying to get work done and it is pausing every 20 minutes or so. That may not seem like much of an intrusion but by about the 4th or 5th time, waiting 10 seconds for it to wake up and come back when you are hot on making some change, it can seem like forever.

    BTW: In every case -- every case -- I can tell when it is going to begin going into the pause cycling: It starts by first becoming unresponsive to mouse input such that I have to click in a spot twice before the text cursor moves to where the mouse cursor is located. Prior to that, I never experience any pausing. After that click-twice behavior occurs, I know it will soon start pausing on me and continue doing so.
     
    Last edited: Jun 9, 2017
    PragmaticPraxis likes this.
  17. PragmaticPraxis

    PragmaticPraxis Bit poster

    Messages:
    74
    @Dmitry@Parallels Any word on this, or has Parallels just given up? Added info, it happened to me again just now, on a VM in which Visual Studio isn't installed, and NotePad++ isn't currently running. The only programs running are Revit 2017 and BlueBeam Revu. The VM is totally frozen, no mouse input, no keyboard input. Work lost. CPU use is low, memory pressure is green. Everything in Mac OS is fine. ONLY Parallels is acting up.
    And, since Parallels has now cost me an hour of work, perhaps I should bill you for that hour? If I caused someone a loss of work I would be responsible for that loss. That's what being a "professional" means. Time to look into EU laws on this, since Parallels can't/won't solve the problem.
    Or perhaps we need to start discussing a class action lawsuit? That Parallels calls this software "Professional" is false advertising, to put it mildly.

    And, twice in one day it completely freezes. Parallels "support" claims "we're working on it" but after three months there's nothing to show for it. Funny thing is, today I got pinged on a Revit forum asking what VM software I use. I said Parallels, but it's a piece of junk now, and after nearly 10 years of use I'm looking at VMWare. Fusion for Mac downloading now. Looks like rather than working this weekend I'll be playing with VMWare. Well done Parallels team. That's called an own goal.
     
    Last edited: Jun 9, 2017
  18. PragmaticPraxis

    PragmaticPraxis Bit poster

    Messages:
    74
    RodB,
    have you tried reducing the number of CPUs assigned to the VM yet? I was using 4, and I froze completely with loss of work (couldn't save in Revit) twice today. But since changing to 2 CPUs I have not frozen. Anecdotal, but maybe worth a try. I think if it keeps working it may put my VMWare transition off by a week or two, but I am pretty much done with Parallels. I just can't trust on it any more.

    Scratch that. It took longer today, but even with the reduced CPU count I am again frozen up. Doubly frustrating, I added the keyboard shortcut I need to save my work to the Parallels customized keyboard shortcuts, in the hopes that I could then send it via the menu, and at least not lose any work. Well, the customization shows up when I edit customizations, but not in the menu. So, yeah, basically @%#&ing useless. What i wouldn't give for Apple to make their own Hypervisor and client that doesn't suck, and include it in a MacOS Pro.
     
    Last edited: Jun 10, 2017
  19. RodB

    RodB Member

    Messages:
    62
    @PragmaticPraxis - My VM is running 4 CPU, 8GB ram, HD has disabled "Real time virtual disk optimization". @Dmitry@Parallels had suggested the latter but it made no difference; haven't bothered to re-enable it.

    However, I noticed something new: I see an option under "CPU & Memory". Under "Advanced Settings" there is a "Hypervisor" dropdown that allows selecting "Parallels" or "Apple". I don't remember seeing that before. Maybe since latest upgrade on Parallels or due to upgrade to Sierra (macOS 10.12.4)? I'll try switching this and see if it makes any difference.
     
  20. Dmitry@Parallels

    Dmitry@Parallels Parallels Team

    Messages:
    702
    Gentlemen,

    My apologies for the delay in the investigation.
    @PragmaticPraxis, according to the technical report your virtual machine is located on the external USB 2.0 disk (WD Passport). If yes, then we strongly recommend you move the VM to the internal Mac disk. USB 2.0 I/O bandwidth is not enough to handle VM disk operations properly.

    @RodB, The Apple hypervisor was indeed added in Parallels Desktop version 12 as an optional (I would say even experimental) hypervisor driver. Apple hypervisor comes short of the following matters comparing to the Parallels hypervisor:
    - Performance: slower on VM startup and shutdown
    - Stability: may crash more frequently
    - Functionality loss: no PMU, nested virtualization, thermal monitoring, energy profiling
    Parallels Hypervisor is the best one.

    Please try to reduce the number of CPUs to 2 and set 4096MB of vRAM to the virtual machine. Please let me know if this helps.
     

Share This Page