I'm running BOINC 7.24.2 on macOS 14.2/Intel.

The suspend while computer is in use isn't working properly. When there is recent keyboard activity, most, but not all, compute tasks suspend.

As I'm typing this message on the machine, most tasks show "Suspended - computer is in use" as the status; however, one task still lists "Running" as its status, and the OS Activity monitor shows the process is using about 100% of one CPU.

The project that is running is from the Rosetta@home project and is listed as Rosetta 4.20 in the application column of the BOINC manager tasks tab. I have one other task listed with that same project and application name that shows as suspended due to the computer being in use. However, the task that is suspended appears to be one that hasn't yet started (0.000% progress), so that may be relevant.

The current task list includes tasks from Einstein@Home, PrimeGrid, and Rosetta@Home. (I'm subscribed to a few more projects, but this system doesn't currently have work units from other projects.). All of the tasks from the other projects correctly show that they are suspended due to computer use. All of the tasks from Einstein@Home and PrimeGrid have started computation. (Progress of at least 5%)

Is this a BOINC problem or a Rosetta@Home problem? How do I fix this?

Same problem here. I'm only running world community grid. I tried using web prefs from http://www.worldcommunitygrid.org/ and tasks ran while the machine was in use. I then tried using local prefs and the same thing happens.

Not only does Bionc fail to suspend while the computer is in use, tasks are running at 100% when my prefs are set to 20%.

Windows 11
Boinc manager 7.24.1

Tail of logging shows work is suspended. But Windows Task Manager clearly shows 16 world community grid threads actively running. All 16 CPU cores show 100% usage.

2024-02-15 18:29:47 | World Community Grid | No tasks sent
2024-02-15 18:29:47 | World Community Grid | No tasks are available for OpenPandemics - COVID 19
2024-02-15 18:29:47 | World Community Grid | No tasks are available for OpenPandemics - COVID-19 - GPU
2024-02-15 18:29:47 | World Community Grid | No tasks are available for Africa Rainfall Project
2024-02-15 18:29:47 | World Community Grid | No tasks are available for Help Stop TB
2024-02-15 18:29:47 | World Community Grid | No tasks are available for Mapping Cancer Markers
2024-02-15 18:29:47 | World Community Grid | No tasks are available for Smash Childhood Cancer
2024-02-15 18:29:47 | World Community Grid | Tasks for AMD/ATI GPU are available, but your preferences are set to not accept them
2024-02-15 18:29:47 | World Community Grid | Tasks for Intel GPU are available, but your preferences are set to not accept them
2024-02-15 18:29:47 | World Community Grid | Project requested delay of 121 seconds
2024-02-15 18:54:53 |  | Windows is suspending operations
2024-02-15 18:54:53 |  | Suspending computation - requested by operating system
2024-02-15 18:54:53 |  | Suspending network activity - requested by operating system
2024-02-15 20:58:19 |  | Resuming after OS suspension
2024-02-15 20:58:19 |  | Resuming network activity
2024-02-15 20:58:26 |  | Windows is resuming operations
2024-02-15 20:58:26 |  | Suspending computation - computer is in use
2024-02-15 20:58:26 |  | Suspending network activity - computer is in use

I too am experiencing this issue. I am only running World Community Grid. This started about 2-3 updates ago. Nothing will suspend. I have triple-checked my settings and nothing seems to resolve the issue.

I am running an iMac Pro (2.5 GHz 14-core Xeon W and Radeon Pro Vega 64 16GB VRAM)) with 64GB RAM and 4TB of HD space. macOS Sonoma 14.3.1.

I have a screen shot but am unsure how to add it to this post. BOINC version 7.24.3 for macOS.

Very frustrating. I have to quit the process to regain control of my computer and often forget to restart it.

I have a screen shot but am unsure how to add it to this post.

As far as I know, to the best of my knowledge...

You would first need to upload the screenshot to a hosting website (Google Photos, Photobucket, etc.), then link to it in a post using the [img] BBC code.

---

https://boinc.berkeley.edu/dl/

I unfortunately had to remove BOINC from my main computer. It had a habit of unsuspending itself during conference calls or other high CPU/GPU tasks and buried the performance of my machine. I'll watch for a release in the future that addresses the issue and consider reinstalling. This was killing my productivity.

You just weren't using the BOINC Manager and client correctly. Boinc is rather difficult for beginners to use and navigate all of its idiosyncrasies,

If you had suspended either the project or the tasks in the Manager, Boinc would have been truly suspended and not interfered with your actual work.

The Boinc gadget in the task bar or whatever Windows thingy is called only suspends Boinc and tasks for an hour and then resumes processing.

If you had suspended either the project or the tasks in the Manager, Boinc would have been truly suspended and not interfered with your actual work.

This does not work either. I have suspended both using the icon in the task bar and also using the drop down menus in the graphical app, and also by changing the app settings. None of these will prevent the app from running tasks.

What OS, what system, what version?

---

Win11 x64-Asus AMD 2.7GHzX4-16GB-GT730 2GB
Garuda x64-ASUS-Intel 2.7GHzX2-16GB-GT450
Mint21.3 x64-Xeon 2.4GHzX4-32GB-nVidia GT730 2GB
Garuda Linux x64-Intel-2.1GHz-8GB
OMV-7 x64-NASIntel-i3-3.6GHz-32GB-29TB

I am having the same problem on Windows 11 Pro. Most of the time Suspend works properly but sometimes it does not. One or more of the Boinc tasks will run full speed, 100% on all cpus and won't stop unless I kill the task or tasks or Boinc. I have noticed this on projects MilkyWay@home and Einstein@home. I am running Boinc 7.24.1 (x64). I have been running Boinc for years and this started only a short while ago.

This is not a production machine so one time I let the task run (It was at 89% completed) and when it finished Boinc went back to its normal behaviour. I wondered if it was the specific project that was broken but now I have seen it happen on different projects.

I wondered if it was the specific project that was broken but now I have seen it happen on different projects.

I my self have not seen this problem on Windows or Linux. More than likely it is PROJECT SPECIFIC.

Best to report it to Milkyway@home as an APPLICATION SPECIFIC BUG if you encounter it again. When reporting the bug to them, try to include as much information as possible.

This has happened with more than one application, Milkyway@home and Einstein@home.

In the Milkyway@home case the Boinc Tasks view showed multiple Milkyway@home tasks as "Suspended - computer is in use (16 cpus)" or "Ready to report (16 cpus)" but one Milkyway@home task showed "Running (16 cpus)". That running task continued and would not stop. If I selected Snooze from the Boinc icon then all Milkway@home tasks changed to "Suspended - user request (16 cpus)" except the running task, which remained at "Running (16 cpus)".

In the Einstein@home case the Boinc Tasks view showed multiple Einstein@home tasks as "Running (0.9 cpus + 1 NVIDIA GPU)" and they would not stop for Snooze.

This problem does not occur every day, but has occurred several times over the past few weeks.

Looking at the Applications page for Milkyway (https://milkyway.cs.rpi.edu/milkyway/apps.php), they currently have just two programs running, and both are multithreaded: one "with Orbit Fitting" and the other without.

You need to identify which variant is refusing to suspend, and notify the Milkyway@home project of that extra detail.

The most BOINC can do is to send a message to the application saying that a suspension has been requested: it's up to the project's programmer for the application in question to ensure that every application receives and responds to these messages.

I also see this in Linux running 7.24.1. It only started after updating to 7.24.1 on April 20. While it was running 7.18.1 it never occurred. I have logs going back to June 2023.

It suspends for 20 seconds and then resumes. The logs show it occurs after an upload to WCG.
I have trouble believing an AMD 3600X can't upload a file and keep on computing.
My lowly laptops running Win10 on a duo-core can do this. But they are running 7.20.2

I can predict when it will occur by looking at the time left on a task in Boincmgr.
So I actually sat at the console at the proper time with Boincmgr/log running and top running.
I see the system activity drop to nothing on the GUI, the tasks drop from top then all comes back after 20 seconds.

Two other machines use the same OS version and client version.
The differences are CPU and memory. The other two both have 32GB and one is a 3900X and the other is a 5950X. So the suspect is memory Right?... except it never occurred with 7.18.1

I suspect the client.

BOINC client version 7.24.1 for x86_64-pc-linux-gnu
#CPUS: 12
CPU model: AMD Ryzen 5 3600X 6-Core Processor [Family 23 Model 113 Stepping 0]
CPU FP OPS: 6912351272.639730
CPU int OPS: 112612226337.385269
OS version: Ubuntu 22.04.4 LTS [5.15.0-105-generic|libc 2.35]
mem size: 16666820608.000000
cache size: 524288.000000
swap size: 2147479552.000000
disk size: 244465692672.000000
disk free: 209108668416.000000


2024-05-13 12:03:52 | World Community Grid | Computation for task MCM1_0216905_4207_0 finished
2024-05-13 12:03:52 | World Community Grid | Starting task MCM1_0216903_4343_0
2024-05-13 12:03:54 | World Community Grid | Started upload of MCM1_0216905_4207_0_r1447428927_0
2024-05-13 12:03:56 | World Community Grid | Finished upload of MCM1_0216905_4207_0_r1447428927_0 (1090 bytes)
2024-05-13 12:03:57 | | Suspending computation - CPU is busy
2024-05-13 12:04:18 | | Resuming computation
2024-05-13 12:32:30 | World Community Grid | Computation for task MCM1_0216903_4337_0 finished
2024-05-13 12:32:30 | World Community Grid | Starting task MCM1_0216905_4166_0
2024-05-13 12:32:32 | World Community Grid | Started upload of MCM1_0216903_4337_0_r1147980307_0
2024-05-13 12:32:34 | World Community Grid | Finished upload of MCM1_0216903_4337_0_r1147980307_0 (1282 bytes)
2024-05-13 12:32:35 | | Suspending computation - CPU is busy
2024-05-13 12:32:55 | | Resuming computation

top - 13:23:41 up 8 days, 16:03, 2 users, load average: 15.34, 15.09, 14.71
Tasks: 269 total, 13 running, 256 sleeping, 0 stopped, 0 zombie
%Cpu(s): 0.0 us, 0.1 sy, 99.9 ni, 0.0 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st
MiB Mem : 15894.7 total, 11822.9 free, 1136.9 used, 2935.0 buff/cache
MiB Swap: 2048.0 total, 2048.0 free, 0.0 used. 14427.2 avail Mem

Each of the 12 Boinc tasks uses 0.2% of memory

Changes in 7.24.1
Improve implementation of CPU throttling

Changes in 7.22.1
Fix non-BOINC CPU usage limit [I tried this at default and 90% with no difference]

I forgot to mention that the client uses default settings except for the length of the queues

I'm having the same issue, approximately half of the active tasks continue to run when the computer is in use. This has happened with both Einstein@home and World Community Grid. The first time it happened with WCG I suspended that project and all the other projects worked as they should. So at first I thought it was just WCG that was the problem. Now Einstein@home has started doing it making me think the problem is not the projects but maybe BOINC. I might add that this PC was running Windows 11 on an AMD Ryzen 9 5900X CPU and was upgraded to an AMD Ryzen 9 5950X. The reason I mention this is because I never encountered tasks continuing to run with the 5900X, only after I switched to the 5950X did the issue start. Makes me wonder if the problem could be somehow related to the number of cores and processors.

---

Given sufficient initial velocity, even pigs can fly!

Makes me wonder if the problem could be somehow related to the number of cores and processors. If you look on the project websites do they list your CPU correctly? If not, try manually running Tools>Run cpu benchmarks and see if it makes a difference. (Maybe try doing that even if they are listed correctly?)

Oops, sorry, meant to put the first bit as a quote.

RE: My Message 114028
I solved my problem by reverting back to version 7.18.1

I seem to be having the reverse problem as most are having. My CPU suspends for 20 seconds after an upload to the project website.

Bottom line it is a CPU suspend problem anyway.

I should never have tried to fix something that was not broken by updating the version.

I've had this issue for quit some time now, Years actually. Starting with Linux Mint, now with Garuda, several flavors of Mint. Don't think I've ever seen this issue on Windows.
Sure wish someone would figure this out.

---

Win11 x64-Asus AMD 2.7GHzX4-16GB-GT730 2GB
Garuda x64-ASUS-Intel 2.7GHzX2-16GB-GT450
Mint21.3 x64-Xeon 2.4GHzX4-32GB-nVidia GT730 2GB
Garuda Linux x64-Intel-2.1GHz-8GB
OMV-7 x64-NASIntel-i3-3.6GHz-32GB-29TB

Seems the only way I have it working correctly on any distro is running Boinc in Flatpak

---

Win11 x64-Asus AMD 2.7GHzX4-16GB-GT730 2GB
Garuda x64-ASUS-Intel 2.7GHzX2-16GB-GT450
Mint21.3 x64-Xeon 2.4GHzX4-32GB-nVidia GT730 2GB
Garuda Linux x64-Intel-2.1GHz-8GB
OMV-7 x64-NASIntel-i3-3.6GHz-32GB-29TB