scan rate and "w" output from vmstat

[Georgios Magklaras]

On 03/07/2011 06:05 PM, travel NJ wrote:
> Hi,
>
> I am in the process of converting a script for gathering system info in
> Solaris to be used in Linux.
>
> I noticed RedHat 5.5 does not have the "w" output from vmstat even the
> manpage indicates that it exists
>
> [sysadm at unix ~]$ vmstat 5 5
> procs                      memory      swap          io     system
> cpu
>   r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy id
> wa
>   1  0      0 2492644 124640 1039360    0    0     0     0    0     0  0  1
> 0  0
>   1  0      0 2492644 124640 1039360    0    0     0    14  103   188 25  0
> 75  0
>   1  0      0 2492644 124640 1039360    0    0     0     0  102   191 25  0
> 75  0
>   1  0      0 2492644 124640 1039360    0    0     0    30  108   189 25  0
> 73  2
> excerpt from man
>
> FIELD DESCRIPTIONS
>     Procs
>         r: The number of processes waiting for run time.
>         b: The number of processes in uninterruptable sleep.
>         w: The number of processes swapped out but otherwise runnable.  This
>            field is calculated, but Linux never desperation swaps.
> In addition, it does not have the "sr" as similar to Solaris's vmstat. I do
> know that many of the info are stalled in /proc, but is /proc/vmstat the
> only location I can get the scanrate ? will it be pgscan_kswapd_high?
>
> thanks
The vmstat manpage of RHEL 5.6 does NOT mention the w column. I have 
checked it out:

FIELD DESCRIPTION FOR VM MODE
    Procs
        r: The number of processes waiting for run time.
        b: The number of processes in uninterruptible sleep.

    Memory
        swpd: the amount of virtual memory used.


The w column on Unix land vmstat refers to the number of procs that are 
runnable but swapped out of RAM. vmstat offers only the number of 
bytes/kb/mb swapped.


Redhat recommends 'dstat' as a replacement for vmstat. Do a
yum -y install dstat

on your RHEL box and  then play with something like:
  dstat -vms

If I were you, I would concentrate on r, b and the free swapped space. 
What matters is not so much how many procs are swapped but the rate of 
forking, the b, r and the free swap that you have. This could give you 
more meaningful stats in a cross-platform overloaded system monitoring 
scenario

GM

-- 
-- 
George Magklaras
Senior Systems Engineer/IT Manager
Biotek Center, University of Oslo
EMBnet TMPC Chair

http://folk.uio.no/georgios

Tel: +47 22840535