On Feb 13, 2009, at 12:00 AM, Fajar A. Nugraha wrote:
Are you using PV drivers?
What domU storage backend did you use? file:? tap:aio:? phy:?
Are you using some kind of LVM snapshot?
yes, we use the xen gpl pv drivers.
the storage backend is phy on a lvm volume
this is our config file:
kernel = "/usr/lib/xen/boot/hvmloader"
name = 'g-hlsan9ib'
memory = '4096'
vcpus = '8'
cpu_weight = '4096'
cpus = "1-7,^0"
vif = [ 'ip=172.16.123.109,mac=00:16:3E:
disk = [ 'phy:/dev/domU/d-jho6fmdr,hda,w','phy:/dev/domU/d-
jmishjr5,hdb,w' ,'file:/root/win2003_r2_64.iso,hdc:cdrom,r', ]
device_model = '/usr/' + arch_libdir + '/xen/bin/qemu-dm'
# Enable/disable HVM guest PAE, default=1 (enabled)
# Enable/disable HVM guest ACPI, default=1 (enabled)
we think it is probably because of disk i/o, because top sometimes
wa (waiting on io) during the hang.
What does iostat say at that time? If i/o wait is 100% but i/o
throughput is low, perhaps you're experiencing the symptoms of disk
I don't think it has to do with disk failure, as our raid controller
reports all disks are "healthy" and there are no errors.
Xen-users mailing list