/dev/blog

If the Dom0’s kernel is built with pciback as a module (grep CONFIG_XEN_PCIDEV_BACKEND /boot/config-`uname -r`), using the kernel command-line parameter pciback.hide won’t work.

There is of course more than solution how you can assign hardware to a domU, but the probably easiest one is to to pass the hide parameter to the pciback module in /etc/modprobe.conf:

[...]
 # hide (0000:05:02.0)
 options pciback hide=(0000:05:02.0)
[...]

The xensouce mailing list even tells you how to modify your initrd preloading the pciback driver before all others

# mkinitrd -f --preload=pciback /boot/initrd-$(uname -r).img $(uname -r)

Another (maybe more elegant) solution would be assigning the pci devices in a separate init script:
http://www.bestgrid.org/index.php/Xen:_assigning_PCI_devices_to_a_domain
http://projects.arcs.org.au/trac/systems/wiki/Howto/PciBack

Update: If you want to avoid rebuilding the initial ramdisk manually every time you update your kernel, you may want to put this small script to your /etc/rc.local

/sbin/modprobe pciback
BDF=0000:05:02.0
# Unbind a PCI function from its driver as necessary
[ ! -e /sys/bus/pci/devices/$BDF/driver/unbind ] || \
        echo -n $BDF > /sys/bus/pci/devices/$BDF/driver/unbind
# Add a new slot to the PCI Backend's list
echo -n $BDF > /sys/bus/pci/drivers/pciback/new_slot
# Now that the backend is watching for the slot, bind to it
echo -n $BDF > /sys/bus/pci/drivers/pciback/bind

With the preceding modprobe pciback the module doesn’t have to be preloaded in you initial ramdisk.

Per default, CentOS 5.3 saves a domU’s state when the hypervisor is shut down. In principle this a nice behaviour: The domU will restore its state automagically when the hypervisor comes up again and just continue working. But there are some services (e.g. mldonkey) that don’t like to be restored and refuse to work (i.e. keep in a blocked state forever).

The easiest way of fixing this, is to shut down the domUs gracefully with the hypervisor. In CentOS (as in probably most other distros as well) you can change xen’s shutdown behaviour in /etc/sysconfig/xendomains

[...]
## Type: string
## Default: /var/lib/xen/save
#
# Directory to save running domains to when the system (dom0) is
# shut down. Will also be used to restore domains from if # XENDOMAINS_RESTORE
# is set (see below). Leave empty to disable domain saving on shutdown 
# (e.g. because you rather shut domains down).
# If domain saving does succeed, SHUTDOWN will not be executed.
#
#XENDOMAINS_SAVE=/var/lib/xen/save
XENDOMAINS_SAVE=
[...]

All we have to do is empty the value of the XENDOMAINS_SAVE variable, so SHUTDOWN will be executed instead.

This is an easy one: You just append zeros at the end of the corresponding image file (here 1GB):

# dd if=/dev/zero bs=1024 count=1000000 >> domU.img

You may want to shutdown your domU before changing its image file. After appending the zeros, you can restart the domain and expand its filesystem.

First, we have to delete our current partition and create a bigger one (don’t be afraid, no data will be lost):

# fdisk /dev/xvda

Type m to get a list of all commands:

Command (m for help): m
Command action
   a   toggle a bootable flag
   b   edit bsd disklabel
   c   toggle the dos compatibility flag
   d   delete a partition
   l   list known partition types
   m   print this menu
   n   add a new partition
   o   create a new empty DOS partition table
   p   print the partition table
   q   quit without saving changes
   s   create a new empty Sun disklabel
   t   change a partition's system id
   u   change display/entry units
   v   verify the partition table
   w   write table to disk and exit
   x   extra functionality (experts only)

Let’s print out the partition table and look for the ext3 partition, we want to enlarge (it’s easy here, there is just one partition):

Command (m for help): p

Disk /dev/xvda: 5218 MB, 5218304000 bytes
255 heads, 63 sectors/track, 634 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

    Device Boot      Start         End      Blocks   Id  System
/dev/xvda1   *           1         634     5092573+  83  Linux

Now we delete the old partition (/dev/xvda1) and create a new one with the maximum available size:

Command (m for help): d
Selected partition 1

Command (m for help): n
Command action
   e   extended
   p   primary partition (1-4)
p
Partition number (1-4): 1
First cylinder (1-634, default 1): 
Using default value 1
Last cylinder or +size or +sizeM or +sizeK (1-634, default 634): 
Using default value 634

Our original /dev/xvda1 had the bootable flag (see fdisk -l output), so we must add it to our new /dev/xvda1 again:

Command (m for help): a
Partition number (1-4): 1

Now let’s write our new partition table and exit fdisk:

Command (m for help): w
The partition table has been altered!

Calling ioctl() to re-read partition table.

WARNING: Re-reading the partition table failed with error 16: Device or resource busy.
The kernel still uses the old table.
The new table will be used at the next reboot.
Syncing disks.

After rebooting the system (!) we can online-resize the filesystem with a simple command:

# resize2fs /dev/xvda1 
resize2fs 1.39 (29-May-2006)
Filesystem at /dev/xvda1 is mounted on /; on-line resizing required
Performing an on-line resize of /dev/xvda1 to 1522150 (4k) blocks.
The filesystem on /dev/xvda1 is now 1522150 blocks long.