[SOLVED] Conky: CPU usage different that htop shows me.

robgeek · 06-07-2014, 09:57 PM

Hi!
I'm using Debian Wheezy with OpenBox in a very old pc, Pentium4 3Ghz.
I would like my Conky1.9 shows my the correct usage of my CPU.
So, the part of conkyrc file that does this is "${cpu}". But when i run htop to make a comparation, htop shows me almost twice of usage of my cpu that Conky does!

Which one is correct and how i can show the correct usage?

Thank you!

Screenshot of my htop and conky on the top of screen.(This time he showed a little bit more of cpu usage that Conky, but he always shows me almost twice of usage)

http://postimg.org/image/3ta7hhrpv/

coralfang · 06-07-2014, 11:31 PM

The ${cpu} value is showing the averaged use. If you look at the values in htop, there are two cores, conky is averaging that.

If you want to show each individual core, add to your conky something like this:

Code:

CPU0 ${cpu cpu0}% | CPU1 ${cpu cpu1}% |

which should show almost identical values as seen in htop.

ondoho · 06-08-2014, 11:59 AM

^ +1 to that.

please keep in mind that measuring cpu usage is generally a pseudo-science and the outcome is only a rough approximation.

rokytnji · 06-08-2014, 12:11 PM

Mine is

Code:

harry@biker:~
$ inxi -f
CPU:       Single core Mobile Intel Pentium III - M (-UP-) cache: 512 KB
           clocked at 798 MHz
           CPU Flags: cmov cx8 de fpu fxsr mca mce mmx msr mtrr pae pge pse
           pse36 sep sse tsc vme
harry@biker:~
$ uname -r
3.14.3-dreamlinux-liquorix

Which is way way older so I just live with

Code:

${cpu}%

So as not to freak out over cpu usage.

273 · 06-08-2014, 12:17 PM

Quote:

Originally Posted by coralfang

The ${cpu} value is showing the averaged use. If you look at the values in htop, there are two cores, conky is averaging that.

If you want to show each individual core, add to your conky something like this:

Code:

CPU0 ${cpu cpu0}% | CPU1 ${cpu cpu1}% |

which should show almost identical values as seen in htop.

I think that that CPU0 is actually the average and you would need to use CPU1 and CPU2 in conky to see the individual readings. Certainly that's how it works in my conky.

Code:

CPU0: ${freq_g 1}GHz: ${cpu cpu1}%${goto 320}CPU1:${freq_g 2}GHz: ${cpu cpu2}%${goto 640}CPU2:${freq_g 3}GHz: ${cpu cpu3}%${goto 980}CPU3:${freq_g 4}GHz: ${cpu cpu4}%
CPU4: ${freq_g 5}GHz: ${cpu cpu5}%${goto 320}CPU5:${freq_g 6}GHz: ${cpu cpu6}%${goto 640}CPU6:${freq_g 7}GHz: ${cpu cpu7}%${goto 980}CPU7:${freq_g 8}GHz: ${cpu cpu8}%

robgeek · 06-08-2014, 05:37 PM

Ok, guys!
I understand now.

Thank you for your answers!

robgeek · 06-08-2014, 09:15 PM

Guys. One more thing. If Conky shows me the average of usage, why when i'm compressing a file my system gets extremely slow and Conky shows me 50% or 60% of usage.
I know, probably i'm using for this task 100% of one core and nothing of the other one, but why my system gets so slow? Why he doesn't use the other core?

TobiSGD · 06-09-2014, 04:47 AM

Because the Pentium 4 never had two cores, the only dual-core Pentiums from that era were the Pentium D CPUs.
What you have is a single core CPU with Hyperthreading, the CPU acts as having two virtual cores to be able to use its different units better. In the case of compressing files the CPU is already using all its parts to the limit so that a better utilization of other parts of the CPU is not possible, making your system slow while it appears to be working at only half its power.