CPU usage is around 70%.
But no program is consuming that amount of CPU resources.
What does this mean? I am using slacko 5.5.
Strange CPU usage
-
- Posts: 229
- Joined: Wed 25 Aug 2010, 15:38
- Location: ISM Dhanbad, Jharkhand, India
Strange CPU usage
- Attachments
-
- strange_cpu_usage.png
- (75.54 KiB) Downloaded 266 times
-
- Posts: 229
- Joined: Wed 25 Aug 2010, 15:38
- Location: ISM Dhanbad, Jharkhand, India
It should be as low as possible.Jasper wrote: I hope that without derailing this thread I can ask what prolonged level of CPU % usage, if any, is considered undesirable and/or inadvisable?
High CPU usage is not the issue I am discussing here.
The question is that CPU usage of all processes add up to around 10%. Then how do I get ~70%?
-
- Posts: 229
- Joined: Wed 25 Aug 2010, 15:38
- Location: ISM Dhanbad, Jharkhand, India
The above screenshot is taken after killing the plugin-container.simargl wrote:If I had to guess, I'd say it's flash plugin high CPU usage (plugin-container)
Maybe install lxtask and sort processes by CPU%
Edit: Even with plugin container high CPU usage cannot be explained
With plugin container CPU usage is ~85%. This should be around 30%.
It appears as if there's some hidden process stuck in an infinite loop.
And lxtask is not helping here. It too doesn't show that 'hidden process'.
This is happening first time. Might not be reproducible in next boot.
- Attachments
-
- strange_cpu_usage_via_lxtask.png
- All CPU usages of individual processes add up to 38% but it shows 85%
- (126.34 KiB) Downloaded 411 times
Hi again,
The link below relates to my questions to the Htop Developer about its reported RAM usage.
http://www.murga-linux.com/puppy/viewto ... f0f8eaf731
My regards
The link below relates to my questions to the Htop Developer about its reported RAM usage.
http://www.murga-linux.com/puppy/viewto ... f0f8eaf731
My regards
-
- Posts: 229
- Joined: Wed 25 Aug 2010, 15:38
- Location: ISM Dhanbad, Jharkhand, India
Does this mean that linux kernel was using that amount of CPU?Jasper wrote:Hi again,
The link below relates to my questions to the Htop Developer about its reported RAM usage.
http://www.murga-linux.com/puppy/viewto ... f0f8eaf731
My regards
The problem disappeared after a reboot. But I'm curious what was happening that time.
Interesting problem. I had something similar recently on Slacko 53 - htop showed no processes consuming excessive cpu, but the small graphical icon down in the system tray showed near 100% (sorry I don't know which programme provides/calculates the sys tray icon on Slacko53...).
I did not pay attention to the total cpu usage bargraph shown by htop (haven't noticed that before) - just the sum of the individual programmes, and none of those was above 2%.
I assume that htop is therefore unable to accurately quantify all of the individual contributors to the overall cpu usage - but suspect that it's figure for the overall usage is correct.
In my case the problem was caused by my testing of an audio player programme which continued playing audio after i shut down the programme which "kicked off" the audio stream in the first place.
In fact I ended up with two audio streams overlaying each other.
Maybe there are some "zombie" functions that can occur within the kernel activity but do not get evaluated by htop. (ie: is the kernel able to force hardware to consume cpu cycles without there being a visible cpu process attached to that usage?)
Could your problem be a symptom of an audio or video process that causes an unreported communication/activity between the kernel and the video accelerator or audio processor etc??
Maybe the kernel THINKS it has passed control to the hardware (eliminating any "official" tracking of the process), but in fact the accelerator fails to kick in, and leaves the cpu handling the action??
(sorry if this doesn't make sense...)
I did not pay attention to the total cpu usage bargraph shown by htop (haven't noticed that before) - just the sum of the individual programmes, and none of those was above 2%.
I assume that htop is therefore unable to accurately quantify all of the individual contributors to the overall cpu usage - but suspect that it's figure for the overall usage is correct.
In my case the problem was caused by my testing of an audio player programme which continued playing audio after i shut down the programme which "kicked off" the audio stream in the first place.
In fact I ended up with two audio streams overlaying each other.
Maybe there are some "zombie" functions that can occur within the kernel activity but do not get evaluated by htop. (ie: is the kernel able to force hardware to consume cpu cycles without there being a visible cpu process attached to that usage?)
Could your problem be a symptom of an audio or video process that causes an unreported communication/activity between the kernel and the video accelerator or audio processor etc??
Maybe the kernel THINKS it has passed control to the hardware (eliminating any "official" tracking of the process), but in fact the accelerator fails to kick in, and leaves the cpu handling the action??
(sorry if this doesn't make sense...)