Re: BUG: rcu_sched detected stalls on CPUs

[mitko]

On 2017-01-20 17:19, Steven Rostedt wrote:
> On Fri, 20 Jan 2017 10:43:50 +0200
> mitko@xxxxxxxxxx wrote:
>
> > [1.] One line summary of the problem:
> >
> > rcu_sched detected stalls on CPUs and few minutes server not respond.
>
> Is this reproducible? Or was this a one time ordeal?

It's happened usual once per day and can't be reproducible. If server 
load average is more then
0.70 it's will happened twice per day.

> > [2.] Full description of the problem/report:
> >
> > Load of my server (postgres database) isnt big less then 0.50 and when
> > error occured rcu_sched detected stalls on CPUs
> > server freeze and nothing is work for 3-5 minute.
> > No network, no video signal, no keyboard, no mouse. Nothing is worked.
> > After these few minutes everything continue normal.
> > This usual is happend once per day. When I check in google find a lots
> > of ppl complain of this error, but no solution.
> > Do any one know can help me to resolve it ? I spoke with few friends 
> > and
> > they trying to convince me the problem is in CPU.
> > I did not believe after a 3 years working CPU suddenly stop working
> > correctly, but I might be wrong.
> >
> > [3.] Keywords (i.e., modules, networking, kernel):
> >
> > kernel
> >
> > [4.] Kernel information
> > [4.1.] Kernel version (from /proc/version):
> >
> > Linux version 4.4.38 (root@hive64) (gcc version 5.4.0 (GCC) ) #2 SMP 
> > Sun
> > Dec 11 16:11:02 CST 2016
>
> Have you tried a newer version of the kernel?

No, for the moment I didnt because I can't find new one in slackware 
repos.
My last kernel compile was almost 7 years ago and I do not feel safe 
now, I guess a lot of things have changed.


> -- Steve

Regards,
Mitko