* can the idle value of /proc/stat decrease?
[not found] <1ad49a62-41dd-cdcc-2f8c-b7a2ad67c3b6@huawei.com>
@ 2020-12-28 7:54 ` xuqiang (M)
2021-03-03 16:36 ` Gioh Kim
0 siblings, 1 reply; 2+ messages in thread
From: xuqiang (M) @ 2020-12-28 7:54 UTC (permalink / raw)
To: adobriyan, christian.brauner, michael.weiss, sfr, Andrew Morton,
linux-kernel, linux-fsdevel
Cc: rui.xiang
Our recent test shows that the idle value of /proc/stat can decrease.
Is this an unreported bug? or it has been reported and the solution is
waiting to get merged.
The results of the two readings from /proc/stat are shown as below, the
interval between the two readings is 150 ms:
cat /proc/stat
cpu0 5536 10 14160 4118960 0 0 227128 0 0 0
cat /proc/stat
cpu0 5536 10 14160 4118959 0 0 227143 0 0 0
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: can the idle value of /proc/stat decrease?
2020-12-28 7:54 ` can the idle value of /proc/stat decrease? xuqiang (M)
@ 2021-03-03 16:36 ` Gioh Kim
0 siblings, 0 replies; 2+ messages in thread
From: Gioh Kim @ 2021-03-03 16:36 UTC (permalink / raw)
To: xuqiang (M)
Cc: adobriyan, christian.brauner, michael.weiss, sfr, Andrew Morton,
linux-kernel, linux-fsdevel, rui.xiang
Hi,
We found the same problem on our production servers.
For example, prometheus-node-exporter reported below message:
daemon.info: Feb 8 17:43:12 ps601b-202
prometheus-node-exporter[2506]: level=warn ts=2021-02-08T17:43:12.052Z
caller=cpu_linux.go:247 collector=cpu msg="CPU Idle counter jumped
backwards, possible hotplug event, resetting CPU stats" cpu=78
old_value=1.04782328e+06 new_value=1.04782178e+06
The idle value was changed from 1.04782328e+06 to 1.04782178e+06.
AND the server rebooted after some minutes.
There is no other error in the log files.
I currently suspect the CPU counter decreasing caused the system reboot.
On Mon, Dec 28, 2020 at 8:57 AM xuqiang (M) <xuqiang36@huawei.com> wrote:
>
> Our recent test shows that the idle value of /proc/stat can decrease.
> Is this an unreported bug? or it has been reported and the solution is
> waiting to get merged.
>
> The results of the two readings from /proc/stat are shown as below, the
> interval between the two readings is 150 ms:
>
> cat /proc/stat
> cpu0 5536 10 14160 4118960 0 0 227128 0 0 0
>
> cat /proc/stat
> cpu0 5536 10 14160 4118959 0 0 227143 0 0 0
--
Gioh Kim
Cloud server kernel maintainer
Quality Management (IONOS Cloud)
1&1 IONOS SE | Greifswalder Str. 207 | 10405 Berlin | Germany
Phone: +49 176 26978962
E-mail: gi-oh.kim@cloud.ionos.com | Web: www.ionos.de
Hauptsitz Montabaur, Amtsgericht Montabaur, HRB 24498
Vorstand: Dr. Christian Böing, Hüseyin Dogan, Dr. Martin Endreß,
Hans-Henning Kettler, Arthur Mai, Matthias Steinberg, Achim Weiß
Aufsichtsratsvorsitzender: Markus Kadelke
Member of United Internet
Diese E-Mail kann vertrauliche und/oder gesetzlich geschützte
Informationen enthalten. Wenn Sie nicht der bestimmungsgemäße Adressat
sind oder diese E-Mail irrtümlich erhalten haben, unterrichten Sie
bitte den Absender und vernichten Sie diese E-Mail. Anderen als dem
bestimmungsgemäßen Adressaten ist untersagt, diese E-Mail zu
speichern, weiterzuleiten oder ihren Inhalt auf welche Weise auch
immer zu verwenden.
This e-mail may contain confidential and/or privileged information. If
you are not the intended recipient of this e-mail, you are hereby
notified that saving, distribution or use of the content of this
e-mail in any way is prohibited. If you have received this e-mail in
error, please notify the sender and delete the e-mail.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2021-03-03 19:28 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
[not found] <1ad49a62-41dd-cdcc-2f8c-b7a2ad67c3b6@huawei.com>
2020-12-28 7:54 ` can the idle value of /proc/stat decrease? xuqiang (M)
2021-03-03 16:36 ` Gioh Kim
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).