[ClusterLabs] Antw: notice: throttle_handle_load: High CPU load detected

Ulrich Windl Ulrich.Windl at rz.uni-regensburg.de
Fri May 22 02:36:02 EDT 2015


Hi!

I Linux I/O waits are considered for load (as opposed to other Linuxes) Thus
if your CPUs have >90% I/O wait, you'll see a high load. Did you try pressing
"1" in top to watch "wait"?

Regards,
Ulrich

>>> Kostiantyn Ponomarenko <konstantin.ponomarenko at gmail.com> schrieb am
21.05.2015
um 17:35 in Nachricht
<CAEnTH0dVubWb3ricPXG1=aFQ9oD14gzitUN5Pnu7Ggr8588YAA at mail.gmail.com>:
> Hi,
> 
> I noticed these messages in the log:
> 
> INFO May 15 03:08:08 crmd(8762) B6-4U-207-T notice: throttle_handle_load:
> High CPU load detected: 19.520000
> INFO May 15 03:07:08 crmd(8762) B6-4U-207-T notice: throttle_handle_load:
> High CPU load detected: 20.400000
> INFO May 15 03:06:38 crmd(8762) B6-4U-207-T notice: throttle_handle_load:
> High CPU load detected: 19.959999
> INFO May 15 03:06:08 crmd(8762) B6-4U-207-T notice: throttle_handle_load:
> High CPU load detected: 19.480000
> 
> The PC has 12 cores and there was only about 30% of usage of all cores,
> when the messages started to appear.
> I found "load−threshold" parameter, but there is said that the default
> value is 80%, and I didn't load the PC that much.
> I looked at the source code and there is a detection of cores number.
> 
> So far I cannot understand what is wrong here?
> I know that it is just a notice, but it bothers me as I don't know why it
> happens.
> 
> Thank you,
> Kostya







More information about the Users mailing list