CPU issues on VAIO

[Archie]

Hi. I was able to get a solution on this issue before and I tried searching for that post but couldn't find it anymore. Anyway, I have for about two months now installed 2010.12 and prior to that install, the Sony VAIO FZ-25 dual-core were both running below 60 degrees C.

Well, it's back to the problem of hot CPUs (running around 75-80 degrees) for me and I seem to remember removing something from modprobe.preload but I really can't be sure and remember what. Can anyone help me remember?

Yes, I should have saved that post but I didn't so burn me at stake!

Thanks in advance for any help.

[menotu]

Yes, I should have saved that post but I didn't so burn me at stake! 

If not on the stake at least on your CPU 

[T6]

apart from the mdprobe, have you done maintenace to that machine?  clean the heatsink?

[Archie]

i suppose it's been a while since i sprayed compressed air on that thing. yeah, i better bring it to the shop for maintenance and see if that clears up the excess heat. thanks.

[DeBaas]

Sony VGN-NW21MF/W

from Synaptic check if:
cpufreq
cpufrequtils
powernowd
are instaaled.
Still a bit power hungry but usable.
Have to take a look at the HDD power on/off in battery modus.

[Archie]

Sony VGN-NW21MF/W

from Synaptic check if:
cpufreq
cpufrequtils
powernowd
are instaaled.
Still a bit power hungry but usable.
Have to take a look at the HDD power on/off in battery modus.

All good and well with the test kernel, and I brought the darn thing in for cleaning.

[DeBaas]

Changed the HDD time out in:
/etc/laptop-mode/laptop-mode.conf
from 20 to 60 sec, I went creazy from the continues down/up spinning of the HDD

Code: [Select]

CONTROL_HD_IDLE_TIMEOUT=1


#
# Idle timeout values. (hdparm -S)
# Default is 2 hours on AC (NOLM_HD_IDLE_TIMEOUT_SECONDS=7200) and 20 seconds
# for battery and for AC with laptop mode on.
#
# LM_AC_HD_IDLE_TIMEOUT_SECONDS=20
# LM_BATT_HD_IDLE_TIMEOUT_SECONDS=20
LM_AC_HD_IDLE_TIMEOUT_SECONDS=60
LM_BATT_HD_IDLE_TIMEOUT_SECONDS=60
NOLM_HD_IDLE_TIMEOUT_SECONDS=7200


[AS]

Changed the HDD time out in:
/etc/laptop-mode/laptop-mode.conf
from 20 to 60 sec, I went creazy from the continues down/up spinning of the HDD

Code: [Select]

CONTROL_HD_IDLE_TIMEOUT=1


#
# Idle timeout values. (hdparm -S)
# Default is 2 hours on AC (NOLM_HD_IDLE_TIMEOUT_SECONDS=7200) and 20 seconds
# for battery and for AC with laptop mode on.
#
# LM_AC_HD_IDLE_TIMEOUT_SECONDS=20
# LM_BATT_HD_IDLE_TIMEOUT_SECONDS=20
LM_AC_HD_IDLE_TIMEOUT_SECONDS=60
LM_BATT_HD_IDLE_TIMEOUT_SECONDS=60
NOLM_HD_IDLE_TIMEOUT_SECONDS=7200


to be fully effective, that's would require to change some param in /etc/fstab too ... adding commit=<no seconds>, the default is to flush out data and metadata every 5 seconds for ext3/4 and every 20 secs for ext2 ...

AS

[horusfalcon]

I know it's stale, but I don't see this [SOLVED] yet, and found this:

http://www.pclinuxos.com/forum/index.php/topic,95065.msg807204.html#msg807204

It might help jog your memory, Archie.  (I'm still searching, too...)

Later On,
D

[Archie]

heya horusfalcon,

Wow, you found this thread ... but it got solved somehow but not in the manner that I thought it would.

My solution? I installed AS' testing kernel and my CPU are running at good cycles ... 7-10% when idle ... and even temperature improved!

[AS]

heya horusfalcon,

Wow, you found this thread ... but it got solved somehow but not in the manner that I thought it would.

My solution? I installed AS' testing kernel and my CPU are running at good cycles ... 7-10% when idle ... and even temperature improved!

It should really be much lower when "idle" ... definitely less than 1%, please PM me if you want further investigate this issue.

AS

[Archie]

heya horusfalcon,

Wow, you found this thread ... but it got solved somehow but not in the manner that I thought it would.

My solution? I installed AS' testing kernel and my CPU are running at good cycles ... 7-10% when idle ... and even temperature improved!

It should really be much lower when "idle" ... definitely less than 1%, please PM me if you want further investigate this issue.

AS

We could do it right up here ... be transparent so others may learn from it as well.

Less than 1% .... hmmm I should tell you that I have several processes running in the background like for example Autokey monitoring for any keystrokes... KMail checking POP3 every 10 minutes, etc.

What would you like me to do?

[AS]

install sysstat package then run the following command from a terminal:

Code: [Select]

sar 15 3
exit all running applications, including net_applet and conky/gkrellm if any and repeat the above command.

switch to runlevel 3, thereby using text console, and run again the above command:

Code: [Select]

init 3


Code: [Select]

sar 15 3

the result should help to identify the context of the unusual high load, being the system, the GUI, or some apps ...

[horusfalcon]

heya horusfalcon,

Wow, you found this thread ... but it got solved somehow but not in the manner that I thought it would.

My solution? I installed AS' testing kernel and my CPU are running at good cycles ... 7-10% when idle ... and even temperature improved!

I seem to be getting better at searching the fora as I go along here.  Glad I could be of service.  I'm also glad to see movement toward a solution.  Thanks for your preference to keep the process in view so that others may profit.

Later On,
D

[Archie]

Results from the test here.

So as I continuously run background apps and monitors, naturally I have to sacrifice burning some CPU cycles. But frankly, I think that its akonadi and nepomuk file indexer that are using much of it. Then there's conky and superkarmaba making calls every second or so.

Any advice? Thanks.