On Tue, Sep 13, 2005 at 11:43:10AM -0400, Rik van Riel wrote:
> On Tue, 13 Sep 2005, master@xxxxxxxxxxxxxxx wrote:
>
> > Spoke too soon on the 1.00 load average. Other than that, both dom0 and
> > domU boot and networking works -- yea!
> >
> > [root@teegeeack ~]# uptime&&uname -a
> > 07:47:13 up 12:40, 2 users, load average: 1.00, 1.00, 1.00
> > Linux teegeeack.bradleyland.com 2.6.12-1.1454_FC4xen0 #1 SMP Fri Sep 9
> > 00:19:20 EDT 2005 i686 i686 i386 GNU/Linux
>
> I'm also seeing the load average problem on my system.
Same here. I think the culprit is the kxbwatch kernel thread, which ps
shows as being in the "D" (uninterruptible sleep) state most of the time
on my system. My understanding of load average computation is that "D"
processes count as running (typically, they correspond to a running
process blocked on I/O).
If this is the case, then the load average of 1.00 should be harmless,
though it would be nice to get the xenbus thread to use a normal sleep
so it doesn't skew the load average.
--Michael Vrable
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
Home