Hi all,
Recently, vmstast show crazy big "bi" value even though the system has
no stress. Is this known issue?
Thanks.
% vmstat 1
procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
r b swpd free buff cache si so bi bo in cs us sy id wa st
1 0 0 1133060 151336 479708 0 0 0 0 50 50 0 0 100 0 0
1 0 0 1133060 151336 479708 0 0 0 0 35 30 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 27 24 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 29 26 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 26 22 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 27 24 0 0 100 0 0
1 0 0 1133060 151336 479708 0 0 0 0 45 49 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 62 54 1 0 99 0 0
0 0 0 1133060 151336 479708 0 0 0 0 54 56 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 28 24 0 1 99 0 0
1 0 0 1133060 151336 479708 0 0 0 0 54 54 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 42 38 0 1 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 31 24 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 33 24 0 0 100 0 0
1 0 0 1133060 151336 479708 0 0 0 0 56 58 1 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 157 225 1 0 98 0 0
0 0 0 1133060 151336 479708 0 0 0 0 40 37 0 0 100 0 0
0 0 0 1133060 151336 479708 0 0 0 0 51 47 0 0 100 0 0
1 0 0 1133060 151336 479708 0 0 0 0 51 46 1 0 99 0 0
0 0 0 1133060 151336 479708 0 0 0 0 296 418 2 0 97 0 0
0 0 0 1133060 151336 479708 0 0 0 0 295 419 3 2 95 0 0
0 0 0 1133060 151336 479708 16 0 0 0 313 449 3 1 96 0 0
0 0 0 1132936 151336 479708 0 0 4294967293 0 331 448 5 1 95 0 0
1 0 0 1132936 151336 479708 0 0 0 0 303 445 3 1 96 0 0
0 0 0 1132936 151336 479708 16 0 0 0 308 433 3 1 96 0 0
0 0 0 1132936 151336 479708 0 0 0 0 307 450 4 0 96 0 0
1 0 0 1132936 151344 479704 96 0 0 0 255 222 6 1 93 0 0
1 0 0 1132812 151344 479708 0 0 4294967293 0 397 310 11 1 88 0 0
0 0 0 1132688 151344 479708 0 0 4294967293 0 410 365 9 1 90 0 0
1 0 0 1132812 151344 479708 0 0 3 0 33 32 0 0 100 0 0
0 0 0 1132812 151344 479708 0 0 0 0 31 23 0 0 100 0 0
0 0 0 1132812 151344 479708 0 0 0 0 35 34 0 0 100 0 0
On Thu, 24 Mar 2011 10:52:54 +0900 (JST) KOSAKI Motohiro wrote:
> Hi all,
>
> Recently, vmstast show crazy big "bi" value even though the system has
> no stress. Is this known issue?
>
> Thanks.
underflow? also looks like -3 or -ESRCH.
Adding Jens in case he has any idea about it.
>
> % vmstat 1
> procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
> r b swpd free buff cache si so bi bo in cs us sy id wa st
> 1 0 0 1133060 151336 479708 0 0 0 0 50 50 0 0 100 0 0
> 1 0 0 1133060 151336 479708 0 0 0 0 35 30 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 27 24 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 29 26 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 26 22 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 27 24 0 0 100 0 0
> 1 0 0 1133060 151336 479708 0 0 0 0 45 49 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 62 54 1 0 99 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 54 56 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 28 24 0 1 99 0 0
> 1 0 0 1133060 151336 479708 0 0 0 0 54 54 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 42 38 0 1 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 31 24 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 33 24 0 0 100 0 0
> 1 0 0 1133060 151336 479708 0 0 0 0 56 58 1 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 157 225 1 0 98 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 40 37 0 0 100 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 51 47 0 0 100 0 0
> 1 0 0 1133060 151336 479708 0 0 0 0 51 46 1 0 99 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 296 418 2 0 97 0 0
> 0 0 0 1133060 151336 479708 0 0 0 0 295 419 3 2 95 0 0
> 0 0 0 1133060 151336 479708 16 0 0 0 313 449 3 1 96 0 0
> 0 0 0 1132936 151336 479708 0 0 4294967293 0 331 448 5 1 95 0 0
> 1 0 0 1132936 151336 479708 0 0 0 0 303 445 3 1 96 0 0
> 0 0 0 1132936 151336 479708 16 0 0 0 308 433 3 1 96 0 0
> 0 0 0 1132936 151336 479708 0 0 0 0 307 450 4 0 96 0 0
> 1 0 0 1132936 151344 479704 96 0 0 0 255 222 6 1 93 0 0
> 1 0 0 1132812 151344 479708 0 0 4294967293 0 397 310 11 1 88 0 0
> 0 0 0 1132688 151344 479708 0 0 4294967293 0 410 365 9 1 90 0 0
> 1 0 0 1132812 151344 479708 0 0 3 0 33 32 0 0 100 0 0
> 0 0 0 1132812 151344 479708 0 0 0 0 31 23 0 0 100 0 0
> 0 0 0 1132812 151344 479708 0 0 0 0 35 34 0 0 100 0 0
>
> --
---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***
On 2011-03-24 17:57, Randy Dunlap wrote:
> On Thu, 24 Mar 2011 10:52:54 +0900 (JST) KOSAKI Motohiro wrote:
>
>> Hi all,
>>
>> Recently, vmstast show crazy big "bi" value even though the system has
>> no stress. Is this known issue?
>>
>> Thanks.
>
> underflow? also looks like -3 or -ESRCH.
>
> Adding Jens in case he has any idea about it.
First question, what does 'recently' mean? In other words, in what
kernel did you first notice this behaviour?
--
Jens Axboe
> On 2011-03-24 17:57, Randy Dunlap wrote:
> > On Thu, 24 Mar 2011 10:52:54 +0900 (JST) KOSAKI Motohiro wrote:
> >
> >> Hi all,
> >>
> >> Recently, vmstast show crazy big "bi" value even though the system has
> >> no stress. Is this known issue?
> >>
> >> Thanks.
> >
> > underflow? also looks like -3 or -ESRCH.
> >
> > Adding Jens in case he has any idea about it.
>
> First question, what does 'recently' mean? In other words, in what
> kernel did you first notice this behaviour?
Thans, Randy, Jens.
I dagged awhile and I've found this is -mm specific and known issue.
It was fixed by Hannes. (ref Message-ID: <[email protected]>)