>> Hmm...have you tried [various suggestions]?
> No, to all of those - and all good suggestions (and since this is a
> XEN DomU I can give it just 1 CPU as well...)

Might also be interesting to see whether/how the symptoms change on
real hardware, if you have any such available; with xen in the mix
there are a bunch more possible places bugs could be causing trouble.

>       time cgdconfig ...

(In passing, I always twitch a little, even if only mentally, when I
see that name: what, you're trying to configure Chris Demetriou?)

> and that reports (when I killed it the cgdconfig, in a couple of
> attempts)  3.xx and 5.xx seconds elapsed time, and the same amount of
> system time  and 0 user time.

> [...], that is clearly backwards.

Indeed.  I'm wondering if whatever system/user time sampling is driven
off of is something xen tries to be "smart" about and, as is far too
common when something tries to be smart, botches.  I've occasionally
seen systems which have load averages up in the hundreds but you'd
never know it from interactive response, because that load is all
phantom load from processes that wake up every clock tick, do almost
nothing, and go back to sleep, taking maybe a percent or two of the CPU
- but, when the run queue is sampled every tick, they're always
runnable.  Something related might be happening here at some level.

/~\ The ASCII                             Mouse
\ / Ribbon Campaign
 X  Against HTML                mo...@rodents-montreal.org
/ \ Email!           7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B

Reply via email to