1

I found that my "PCI Adapter" overheats when load is above 4.0 (having 4 cores). It does NOT overheat when the load is below 4.0 (but I had an overclocked memory that was causing trouble as I explain at the end), so my focus is on lowering the system load. What overheats is just the k10temp-pci-00c3 PCI Adapter (according to sensors).

The mistake:
I am trying to track applications that compose that value, and how much each did.

From this question, I found atop, but reading its output, it is not clear what applications composed the system load... It is like a "simple" column showing SYSLOAD is missing. It seems I could use all that displayed information to calculate something that could show such column, but I just don't know how.

The explanations I found about system load seems too generic (I can be wrong..), so I couldn't understand them enough to produce an algorithm; I think I should mix some data like cpu usage, io usage, mem usage and so get the sysload even if it is somewhat fuzzily guessed...

Precise info on how system Load works:
on this question there is a reference to a pdf explaining it. Load seems not exactly related to specific applications...

Still need a workaround:
Anyway I still would like to know how possible it is to make an application approach about system load, may be there is some way to prioritize the whole system utilization to some specific application, while all the others clogging it, even when they are NOT using much cpu (less than 3%), calm down?

Tests on-going:
I set all processes to nice -n 19, but the one window/process I am actively using; I saw improvement on the temperature graph with psensor (the graph is a lot less fuzzy); the temperature is holding longer, below the limit, now; all processes seems to be behaving as expected; the system load is still high but I saw it once go til 6.0 and the temperature was still fine; no sudden shutdowns since.. need more test tho...

Finally the culprit was found:
My memory was supposed to accept 2000MHz. Long ago I set it to work at 1600MHz, because the machine would not boot otherwise. Now I set it to automatic, what made it work at 1333MHz. The temperature k10temp-pci-00c3 that was overheating (reaching above 80c) now is consistent around 60c when I run heavy applications! Also the load average is around 3.5 to 4.0 without causing any hardware problems! The nice 19 on processes helped postpone the overheating indeed, but wasnt able to totally prevent it!

6
  • Can't you just run top and monitor its output?
    – terdon
    Oct 7, 2013 at 16:47
  • thats the problem, I got kill -SIGSTOP ... on a process that was using 130% cpu (4core max would be 400% I think), the cpu load went from 4.2 to 2.8, and in the list of process using cpu it shows several applications using less than 5% cpu.. Xorg using 15%.. my guess is the load is elsewhere, not something I can guess by reading cpu percentuals. I got about 18 xterm running on a single viewport on compiz, each with its script for monitoring and other tasks; one of them had fast/high text output, I modified it and helped on lowering the load; but the others refresh each 5 seconds;what I do? Oct 7, 2013 at 17:04
  • The output of lm-sensors is very system dependent and will often report sensor data with approximate naming. If that is the only temperature reported by sensors it is almost certainly the CPU temperature. On AMD quad-core cpus (among others), the PCI controller has migrated onto the CPU chip from its historically separate "northbridge" chip.
    – msw
    Oct 7, 2013 at 23:24
  • thats it, mine is amd fx4100 Oct 7, 2013 at 23:34
  • 2
    Load is not directly related to system temperature. You can have a load of 1000 and the system be completely idle and cool. Such things can be caused by network, or disk IO (resulting in an unterruptable sleep). Your question should be "how can I prevent the system from overheating".
    – phemmer
    Oct 8, 2013 at 3:19

2 Answers 2

5

You're asking the wrong question: you've got an overheating system which should be solved by cooling the system. Playing games with process load is going to yield an unsatisfying hack. And since you've got hardware running at its thermal limits, you can fairly expect that problem to worsen.

If you cannot remedy the hardware, see if you can slow the whole processor clock down. Yep, you'll get a slower computer but it is the only reliable way to reduce power consumption / heat dissipation.

5
  • I will try the clock trick thx! btw, I use psensor to monitor the cpu temperature; there is a single temperature that gets crazed on high load k10temp-pci-00c3, the machine beeps, and if beeps too long it shuts down.. I dont care what other applications are doing, even if IO, the application that has focus should work at best and the other should calm down; I improved the cooling system, may be not well enough.. but I am still not sure it will help considering the load goes above 4.0 (on quad core)? so I am focusing on lowering the load, even if I get a better cooling system (with peltier). Oct 7, 2013 at 17:10
  • You might also be able to turn off a core or two and not have to reduce the clock speed. A computer that is using half its cores is infinitely faster than one that shut down because of thermal overload.
    – msw
    Oct 7, 2013 at 17:16
  • but my system load gets above 4.0, if I shutdown any, the limit will lower too! I only get the overheat when load goes above 4.0; below it the machine runs fine! thats why I believe even with a better cooling system I may still have problems.. Oct 7, 2013 at 17:27
  • If the load average (that is, number of jobs in the ready-to-run state) goes above 4, the others will just have to wait. You can't use more than 100% of a core even if there are 23 jobs ready to use it.
    – msw
    Oct 7, 2013 at 23:10
  • I just did something interesting; I put all processes to nice 19, less the one process I am currently using; the temperature on psensor graph got much more normal now, it was very unstable (too high/too low); also I saw the system load go to above 6.0 (once only) and the temperature wasnt that high! all processes were working properly as expected also. I think this need more tests to have some better conclusion... Oct 8, 2013 at 2:01
3

System load isn't really directly related to how much work the system is doing. You could have a load average of only 2.0 and be doing a lot more work than a load average of 8.0.

All the load tells you is the average number of programs eligible to be run. If they are all waiting on your overloaded disk, your CPU won't be doing much of anything, but your load average will still be high.

Ultimately, you have a hardware problem (your "PCI adapter" overheats). You will have a tough time controlling temperature by trying to limit how much runs on your CPU. The process schedulers aren't made that way.

4
  • so the load average may not be directly related to the resulting temperature? I could have a high temperature with a load of 2.0 then.. may be, like msw said, I should focus on the temperature and may be, clock down my cpu at the BIOS, as cpufreq is not working when I try to change the clock. Have to try these and see what happens; it is still getting overheated; as I remember long ago, with an old CPU, it NEVER overheated, at most it got slowed down; so something seems really wrong on my machine! Oct 8, 2013 at 2:55
  • 1
    @AquariusPower Yeah, that would probably help, clocking down the CPU. Also, keep in mind, your fans might be getting older, so they might not be moving as much air too. But yeah, load average is one of those numbers that is a very, very rough guide to how busy the machine is. But %cpu usage is more useful for what you are looking at. But ultimately, reducing how much heat gets produced is the answer.
    – kurtm
    Oct 8, 2013 at 3:10
  • thx again! I revamped my old fans! they were blowing almost no air inside the computer... I have a dust protection net on them that lowered their efficiency in about 80% I guess... so they were useless; I stretched the net now the air is flowing greatly! and indeed they are contributing a lot to lower the temperature! Oct 15, 2013 at 20:30
  • @AquariusPower Glad to hear you solved the issues.
    – kurtm
    Oct 15, 2013 at 20:34

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .