Frequent beachball due to WindowServer (PB4)

Hi,


Since upgrading to PB4 (was fine on previous PB), I've started to experience frequent beachballing (anything from 1 per 5 minutes to 1 every 15 seconds). When the beachball appears, the UI stops updating, I can only move the mouse with the beachball cursor. If audio is playing, this does continue playing, however.


From monitoring system.log, it looks like a WindowServer issue:


Aug 9 16:01:14 Kanto watchdogd[335]: [watchdog_daemon] @(_wd_daemon_service_thread) - service (com.apple.WindowServer) reported as unresponsive
Aug 9 16:01:15 Kanto watchdogd[335]: [watchdog_daemon] @(__wd_service_report_unresponsive_block_invoke) - failed to gather a spindump for (com.apple.WindowServer)
Aug 9 16:01:25 Kanto WindowServer[4570]: send_datagram_available_ping: pid 5897 failed to act on a ping it dequeued before timing out.
Aug 9 16:01:25 Kanto WindowServer[4570]: send_datagram_available_ping: pid 5649 failed to act on a ping it dequeued before timing out.
Aug 9 16:01:25 Kanto WindowServer[4570]: send_datagram_available_ping: pid 5543 failed to act on a ping it dequeued before timing out.
Aug 9 16:01:25 Kanto WindowServer[4570]: send_datagram_available_ping: pid 4821 failed to act on a ping it dequeued before timing out.
Aug 9 16:01:25 Kanto WindowServer[4570]: send_datagram_available_ping: pid 4806 failed to act on a ping it dequeued before timing out.
Aug 9 16:01:25 Kanto WindowServer[4570]: send_datagram_available_ping: pid 4796 failed to act on a ping it dequeued before timing out.
Aug 9 16:01:25 Kanto WindowServer[4570]: send_datagram_available_ping: pid 4735 failed to act on a ping it dequeued before timing out.
// many more of the same message
Aug 9 16:01:33 Kanto WindowServer[4570]: Surface testing disallowed updates for 10 sequential attempts...
Aug 9 16:01:33 Kanto WindowServer[4570]: Surface test allowed updates after 10 attempts (40 ms)


I have a Mid 2014 15 inch Retina Macbook Pro (2.2GHz i7, Iris Pro, 16GB RAM)


Any advice?

Answered by caseyWebb in 39850022

I was able to get the issue to subside by disabling "Automatically adjust brightness" and "Ambient light compensation" under System Preferences > Displays

I am seeing the same thing with my MacBook Pro (Retina, 15-inch, Late 2013)


I am puzzled. The problem did not start with installing the latest beta. But started only today.

As far as I can see, it happens whether or not the browser is running.


I noticed something.

It seems to happen when I am at a shared workspace.

It never seems to happen when on my home network.


Could be a coincidence.

I'm also having this issue, although I've noticed it's intermittant. Usually I've noticed that if I leave my laptop alone for an hour or two it will go away.


That said though, it is interesting that I am on litterally the exact same model MBP as you. Is anyone else here on a Mid 2014 15 inch Retina Macbook Pro (2.2GHz i7, Iris Pro, 16GB RAM)?

Experiencing exactly the same thing on a 2015 12" MacBook, generally while Safari is running. The WiFi network isn't a certificate-secured, or Enterprise one, it's just a standard WPA2 Personal.


Haven't noticed a pattern yet (e.g. it froze while typing this line)

Accepted Answer

I was able to get the issue to subside by disabling "Automatically adjust brightness" and "Ambient light compensation" under System Preferences > Displays

Came here looking for an answer to the same problem.


[Minor sidebar: I am on 10.11 beta (15A243d). I call this "public beta 3". Apple released beta 1, a supplemental beta update, beta 2, and then beta 3. I believe many sites are incorrectly labelling this "beta 4". If you go to the App Store app and look at the installed updates list, it shows "Version 3". Does anyone actually see "Version 4"? I do not and no updates are available.]


I was intrigued by those who made reference to enterprise WiFi as I was also suspecting something on the wireless side before I found this thread. However, on that point I had switched from our WPA2 secured Wifi to non-secured from the same location and saw no change in behavior. My thought was to change to an ethernet connection to bypass WiFi, however, I don't have a USB to ethernet adapter for this Air. (Anyone else tried this?)


I am not seeing any of the ping errors that Keybuk was seeing, I suspect that is unrelated. I am getting a spindump paired with every watchdog error log, like:


8/11/15 1:38:55.708 PM spindump[415]: Saved userspace_watchdog_timeout.spin report for WindowServer version ??? (???) to /Library/Logs/DiagnosticReports/WindowServer_2015-08-11-133855_Brians-MacBook-Air.userspace_watchdog_timeout.spin


8/11/15 1:38:55.712 PM watchdogd[348]: [watchdog_daemon] @(__wd_service_report_unresponsive_block_invoke) - spindump gathered for (com.apple.WindowServer) at (/Library/Logs/DiagnosticReports/WindowServer_2015-08-11-133855_Brians-MacBook-Air.userspace_watchdog_timeout.spin)


Each spindump report starts with:


Command: WindowServer

Path: /System/Library/Frameworks/ApplicationServices.framework/Frameworks/CoreGraphics.framework/Resources/WindowServer

Version: ??? (???)

Parent: launchd [1]

PID: 181


Event: service progress watchdog

Duration: 1.00s

Steps: 10 (100ms sampling interval)


Hardware model: MacBookAir6,1

Active cpus: 4


Does anyone know if there is a way to submit these reports back to the development team?

Wow, moments after my post I saw yours. I don't know how you picked up on this, bravo! But this is absolutely on point. More specifically, disabling the Ambient light compensation option, which I also just discovered is a new setting under El Capitan, elimates the pinwheels immediately. Watching the logs, I went from a spin dump every 1-3 minutes to none for over 20 minutes.


I had to laugh to (and at) myself because all of this came about as I sat in a training room at work in front of windows, convincing myself the problem was somehow wireless related. 😁


Now how to file that bug report with Apple...

I can confirm. After disabling this, the beachballing stopped.

Thanks! This has resolved the issue for me.

Frequent beachball due to WindowServer (PB4)
 
 
Q