[Xen-users] DomU Disk-Intense Operations kill Xen-Server?

[bigfoot29@xxxxxxxxxxxxxxxxxxxxxx]

Hi!

I am sorry to bother you once again, but I wonder how I can get around the
Xen-Crash I can reproduce when combining two xen systems using a drbd
device... as soon as the raid is working (means we have an high latency
network), the DomU starves (and takes Dom0 with it) instantly (e.g. dd
if=/dev/zero of=/file-in-domU-fs.img

I am not SURE if its a Xen-only problem, however, I've never seen that
behaviour before.
I am running Xen 3.0.4-1 in 32-Bit Mode atm and messed around with quite
every setting drbd has to offer.
The funny thing is that I can see (xm top) the peak of the cpu load the
domain creates, but then it just dies.
Is there a way around the CPU-Peak? How to enable synchronous writing? (if
possible)
Is there another fix or Filesystem that does handle these sorts of
networks with more ease? Is this (maybe) a xen problem that could be
addressed in another way?

Weird thing is: that only happens when running the "dd" command inside a
domU - it doesn't do ANY harm when invoked directly at dom0's drbd
device's filesystem

So I am sorry to bother you again, but help is highly appreciated :)

Thanks in Advance!

Regards, Bigfoot29

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users