C
CJH_WkP
Hi all - I've recently bought a new machine. Lenovo P17 with XeonW-10855M@2.8GHZ, NVidia Quadro RTX4000 (8GB), 32GB mem, 2 TB in 2 SSDs. I bought it with Win10Pro (workstation) Installed. I use it for Work and Personal use and have two users on the machine - one personal, one work. Thunderbolt docking station connecting two external monitors and external sound card.
I bought this as an upgrade to a 5 year old P70 on which I was running all the same programs I've installed on the new machine.
For work, the machine has Microsoft365 installed on the work user under one license, and for personal use, it has MSFT365 installed on another license.
Overall, the machine runs very well - handles all my work stuff just fine.
On the personal side, I game on the the machine and run music production programs. (This should be noted now - the game I play and the music apps all run on the OLD machine, just a little slow).
After getting the machine - I updated all drivers, updated windows, then started installing my software packages. All went well and snappy. However, when I started gaming on the machine It was fantastic - great response, superb FPS (well over 100) and it was much improved over my old machine. Same with the music applications. But after a while of using the machine, it would suddenly become sluggish and FPS would crater to less than 30 and the music production software would fail to be even able to play sounds. In fact, windows itself became laggy - opening folders would be slow, file directories would take a long time to populate etc. Meanwhile, my CPU and RAM usage were all very low. No obvious memory hog programs or anything of that nature.
I started a process to ID the problem and see if it was something I did I could fix. Here's a run down:
1.) Complete Fresh install of windows from Sys recovery (twice).
2.) Several runs of DISM /RestoreHealth and sfc / scannow
3.) One by one testing of apps after safe mode reboots to ID any particular problems
4.) Update of all drivers and selective booting with limited devices attached. (ie - I've tried it all the way down to just the laptop and added back peripherals one at a time.).
5.) Updated Bios
6.) Learned how to use the "Ultimate Performance Mode" through Power Options in Lenovo Vantage
7.) Setup and tried the "Ultimate Performance" power plan through the power shell script
8.) Installed / Uninstalled GPU drivers / software managers
9.) Tried switching the BIOS graphics setting to Discrete to only use the RX4000
10.) Installed Advanced System Care and Driver Booster to double check win registry problems / Driver issues
11.) Actively Monitored performance with Sysinternals / ProcMon ProcExp
12.) Reviewed Windows Events with EventViewer and Task Manager to try to ID the offending program.
13.) Set a Group Policy to turn off Power Throttling
14.) Booted without Lenovo Power Manager being active.
Here's where I am: When I reboot the machine - It runs great, for 20 Mins. The, the FPS crater and the audio production gets laggy. I found that the performance degrades at exactly the same time that the Event Viewer Records a series of windows errors related to the WMI. In particular, they are a series of 5858 errors with the details related to a "Operation = Start IWbemServices::ExecQuery" where the result was 0x80041032 (cancelled). Sometimes the "PossibleCause" is listed as relating to "Throttling Idle Tasks, refer to CIMOM regkey: ArbTaskMaxIdle." These errors occur at exactly the same time the machine's performance degrades.
I found these:
New Arbiter Behavior
Other users with same problem2
Other Users with same problem1
Note about the exact error with Resolution
After reading, I created a created custom view in EventViewer and logged WMI errors and started the process of tracing the Client IDs back to the active Tasks to see what programs were creating the problems. There were several that were the offenders and they constantly changed. Sometimes they are programs related to my mouse, sometimes, they were related to the graphics drivers, sometimes they were related to other windows system programs. It never seemed to only be from one or two programs that I could just "shut down." I'd also point out that when I did a safe boot with almost no programs or services running other than system, it still recorded these errors in the Event Viewer. Also note that its is noted in the article about the new arbiter behavior responds to tasks at the 20 minute time mark by default.
Now - a "cancelled" query - which is what a 0x80041032 error means in a 5858 event doesn't sound like it should cripple a system. However, they are occurring at exactly the time when the machine's performance drops.
In particular - here's what happens when I'm gaming. When I first load a game - the system responds nicely - the CPU and GPU both experience and increase in usage, frequency and temperature. The fans on the unit respond and the internal temps get to about 70C degrees. (I've looked it up and that's well below the 90Cdegrees where the GPU would be overheated). Everything is great - game is beautiful and FPS are steady at between 125 and 144. Then, right at the 20 minute mark, those errors are recorded, and my GPU will now still be operating at about 97% capacity, but its temperature will drop dramatically, the fans all slow down and the crappy performance will continue until I reboot the machine. The GPU has never gotten hotter than 75C degrees. Importantly, from that point forward, the GPU will not again experience any increase in temperature until I reboot.
Then, whether I'm gaming or not - those errors are recorded at 20 mins after I reboot and the performance will be terrible. Even though the GPU never heated up (in other words - the GPU is not "overheating" and getting throttled temporarily. It appears to be getting throttled either because of those errors or because of the underlying condition that also created those errors.
If I read the MSFT documentation correctly - the new arbiter behavior just pushed out into Windows Server 2012 but I'm on Win10Pro. I'm guessing W10Pro may be partially based on the WinServer and may have the same behavior implemented. In the article with the "resolution" it suggests that application developers need to change their WMI query from the "ExecQuery" language to the ":: Next" language to deal with the new Arbiter behavior.
In the meantime, I have a crippled machine 20 mins after it boots up. Has anyone seen a way for this to be fixed or repaired? Am I really at the mercy of waiting for application developers to change their software to deal with the new Arbiter? Have I overlooked something I can do to make the machine not have this problem? Any guidance or thoughts would be appreciated. I'm sure you can all imagine how frustrating it is to buy expensive new hardware only have errors being thrown by what seems to be Win10 that are making it unusable.
Continue reading...
I bought this as an upgrade to a 5 year old P70 on which I was running all the same programs I've installed on the new machine.
For work, the machine has Microsoft365 installed on the work user under one license, and for personal use, it has MSFT365 installed on another license.
Overall, the machine runs very well - handles all my work stuff just fine.
On the personal side, I game on the the machine and run music production programs. (This should be noted now - the game I play and the music apps all run on the OLD machine, just a little slow).
After getting the machine - I updated all drivers, updated windows, then started installing my software packages. All went well and snappy. However, when I started gaming on the machine It was fantastic - great response, superb FPS (well over 100) and it was much improved over my old machine. Same with the music applications. But after a while of using the machine, it would suddenly become sluggish and FPS would crater to less than 30 and the music production software would fail to be even able to play sounds. In fact, windows itself became laggy - opening folders would be slow, file directories would take a long time to populate etc. Meanwhile, my CPU and RAM usage were all very low. No obvious memory hog programs or anything of that nature.
I started a process to ID the problem and see if it was something I did I could fix. Here's a run down:
1.) Complete Fresh install of windows from Sys recovery (twice).
2.) Several runs of DISM /RestoreHealth and sfc / scannow
3.) One by one testing of apps after safe mode reboots to ID any particular problems
4.) Update of all drivers and selective booting with limited devices attached. (ie - I've tried it all the way down to just the laptop and added back peripherals one at a time.).
5.) Updated Bios
6.) Learned how to use the "Ultimate Performance Mode" through Power Options in Lenovo Vantage
7.) Setup and tried the "Ultimate Performance" power plan through the power shell script
8.) Installed / Uninstalled GPU drivers / software managers
9.) Tried switching the BIOS graphics setting to Discrete to only use the RX4000
10.) Installed Advanced System Care and Driver Booster to double check win registry problems / Driver issues
11.) Actively Monitored performance with Sysinternals / ProcMon ProcExp
12.) Reviewed Windows Events with EventViewer and Task Manager to try to ID the offending program.
13.) Set a Group Policy to turn off Power Throttling
14.) Booted without Lenovo Power Manager being active.
Here's where I am: When I reboot the machine - It runs great, for 20 Mins. The, the FPS crater and the audio production gets laggy. I found that the performance degrades at exactly the same time that the Event Viewer Records a series of windows errors related to the WMI. In particular, they are a series of 5858 errors with the details related to a "Operation = Start IWbemServices::ExecQuery" where the result was 0x80041032 (cancelled). Sometimes the "PossibleCause" is listed as relating to "Throttling Idle Tasks, refer to CIMOM regkey: ArbTaskMaxIdle." These errors occur at exactly the same time the machine's performance degrades.
I found these:
New Arbiter Behavior
Other users with same problem2
Other Users with same problem1
Note about the exact error with Resolution
After reading, I created a created custom view in EventViewer and logged WMI errors and started the process of tracing the Client IDs back to the active Tasks to see what programs were creating the problems. There were several that were the offenders and they constantly changed. Sometimes they are programs related to my mouse, sometimes, they were related to the graphics drivers, sometimes they were related to other windows system programs. It never seemed to only be from one or two programs that I could just "shut down." I'd also point out that when I did a safe boot with almost no programs or services running other than system, it still recorded these errors in the Event Viewer. Also note that its is noted in the article about the new arbiter behavior responds to tasks at the 20 minute time mark by default.
Now - a "cancelled" query - which is what a 0x80041032 error means in a 5858 event doesn't sound like it should cripple a system. However, they are occurring at exactly the time when the machine's performance drops.
In particular - here's what happens when I'm gaming. When I first load a game - the system responds nicely - the CPU and GPU both experience and increase in usage, frequency and temperature. The fans on the unit respond and the internal temps get to about 70C degrees. (I've looked it up and that's well below the 90Cdegrees where the GPU would be overheated). Everything is great - game is beautiful and FPS are steady at between 125 and 144. Then, right at the 20 minute mark, those errors are recorded, and my GPU will now still be operating at about 97% capacity, but its temperature will drop dramatically, the fans all slow down and the crappy performance will continue until I reboot the machine. The GPU has never gotten hotter than 75C degrees. Importantly, from that point forward, the GPU will not again experience any increase in temperature until I reboot.
Then, whether I'm gaming or not - those errors are recorded at 20 mins after I reboot and the performance will be terrible. Even though the GPU never heated up (in other words - the GPU is not "overheating" and getting throttled temporarily. It appears to be getting throttled either because of those errors or because of the underlying condition that also created those errors.
If I read the MSFT documentation correctly - the new arbiter behavior just pushed out into Windows Server 2012 but I'm on Win10Pro. I'm guessing W10Pro may be partially based on the WinServer and may have the same behavior implemented. In the article with the "resolution" it suggests that application developers need to change their WMI query from the "ExecQuery" language to the ":: Next" language to deal with the new Arbiter behavior.
In the meantime, I have a crippled machine 20 mins after it boots up. Has anyone seen a way for this to be fixed or repaired? Am I really at the mercy of waiting for application developers to change their software to deal with the new Arbiter? Have I overlooked something I can do to make the machine not have this problem? Any guidance or thoughts would be appreciated. I'm sure you can all imagine how frustrating it is to buy expensive new hardware only have errors being thrown by what seems to be Win10 that are making it unusable.
Continue reading...