Re: Load average with CURRENT

On Sun, 7 Mar 2004, Justin Dossey wrote:

> Strangely, I'm seeing doubled load average numbers for my CURRENT
> build.  This machine runs seti_at_home, so it should show a LA around 1
> all the time.  After upgrading from 5.1-RELEASE to 5.2-CURRENT, LA
> doubled.
>
> > uname -a
> FreeBSD shady.home 5.2-CURRENT FreeBSD 5.2-CURRENT #0:
> Fri Mar  5 22:14:05 PST 2004
> jbd_at_shady.home:/usr/obj/usr/src/sys/CUSTOM  i386
> > w
> 10:39AM  up 1 day, 12:06, 1 user, load averages: 2.10, 2.03, 1.94
> USER             TTY      FROM              LOGIN_at_  IDLE WHAT
> jbd              p0       stark            10:39AM     - w
> > vmstat 2 5
>  procs      memory      page                    disks     faults
> cpu
>  r b w     avm    fre  flt  re  pi  po  fr  sr ad0 fd0   in   sy  cs us sy id
>  2 0 0   47780 227080  284   0   0   0 285   0   0   0  369 3213 317 97  3  0
>  1 0 0   47912 226948  372   0   0   0 356   0   0   0  380  122 328 95  5  0
>  1 0 0   47912 226948    0   0   0   0   0   0   0   0  380 10164 327 97  3  0
>  1 0 0   47912 226948    0   0   0   0   0   0   0   0  380 70483 324 68 32  0
>  1 0 0   48808 226048  386   0   0   0 275   0   0   0  379 1505 322 93  7  0
> > sysctl vm.loadavg
> vm.loadavg: { 2.03 2.04 1.97 }
>
> I'll update to the newest CURRENT soon, but I haven't seen any
> discussion of this on the list or elsewhere.  Ideas?

Is this a system with multiple CPUs or a P4 with hyperthreading? It could
be that you didn't previously have the APIC and SMP (or hyperthreading)
options enabled in your kernel and now you do... Could you paste the first
15 lines of the output of "top -S"?

Regards,
Andy

> Andre Guibert de Bruet | Enterprise Software Consultant >
> Silicon Landmark, LLC. | http://siliconlandmark.com/    >