Startup Menu -> Boot Menu -> Security -> System Security
Virtualization Technology (VTx) -> Enabled
Virtualization Technology Directed I/O (VTd) -> Enabled
Both Servers
Choose 'Use entire disk with LVM'
Network setup
apt-get install bridge-utils
vi /etc/network/interfaces
#
auto xenbr0
iface xenbr0 inet dhcp
bridge_ports eth0
auto eth0
iface eth0 inet manual
#
vi /etc/hosts
#
147.8.177.50 cheng-HP-Compaq-Elite-8300-SFF
147.8.179.243 wang-HP-Compaq-Elite-8300-SFF
#
Server #1
Short-cut install for dependencies
apt-get build-dep xen
apt-get install libc6-dev libglib2.0-dev libyajl-dev yajl-tools libbz2-dev bison flex zlib1g-dev git-core texinfo debhelper debconf-utils debootstrap fakeroot
Server #1
apt-get install transfig libpixman-1-dev
#latest
apt-get install git-core
git clone git://xenbits.xen.org/xen.git
#release
wget http://bits.xensource.com/oss-xen/release/4.2.1/xen-4.2.1.tar.gz
tar zxvf xen-4.2.1.tar.gz
cd xen-4.2.1
wget http://download.locatrix.com/xen/remus-qdisc-py.patch
patch -p1 < remus-qdisc-py.patch
./configure
vi ./.config
#
PYTHON_PREFIX_ARG=--install-layout=deb
#
make world
make deb
tar zcvf xen-dist.tgz ./xen-4.2.1/dist
# scp to server #2
Server #2
tar zxvf xen-dist.tgz
Both servers
cd /home/eross/xen-4.2.1/dist
./install.sh
update-grub
ls -al /boot/xen*
-rw-r--r-- 1 root root 802303 Jan 30 20:57 /boot/xen-4.2.1.gz
lrwxrwxrwx 1 root root 12 Jan 30 20:57 /boot/xen-4.2.gz -> xen-4.2.1.gz
lrwxrwxrwx 1 root root 12 Jan 30 20:57 /boot/xen-4.gz -> xen-4.2.1.gz
lrwxrwxrwx 1 root root 12 Jan 30 20:57 /boot/xen.gz -> xen-4.2.1.gz
-rw-r--r-- 1 root root 15388780 Jan 30 20:57 /boot/xen-syms-4.2.1
update-rc.d xencommons defaults 19 18
update-rc.d xendomains defaults 21 20
update-rc.d xen-watchdog defaults 22 23
grep Xen /boot/grub/grub.cfg
# You should see "Ubuntu GNU\/Linux, with Xen hypervisor" in there, that's where I got the below from
sed -i 's/GRUB_DEFAULT=.*\+/GRUB_DEFAULT="Ubuntu GNU\/Linux, with Xen hypervisor"/' /etc/default/grub
update-grub
reboot
Server #1
apt-get install autoconf build-essential
wget http://remusha.wikidot.com/local--files/configuring-and-installing-remus/drbd-8.3.11-remus.tar.gz
cd ./drbd-8.3.11
./autogen.sh
./configure --prefix=/usr --localstatedir=/var --sysconfdir=/etc --with-km
make
make install
cd drbd
make clean all
make install
vi /etc/modules
# add drbd
#
tar zcvf drbd.tgz ./drbd-8.3.11
# scp to server #2
Server #2
tar zxvf ./drbd.tgz
cd ./drbd-8.3.11
make install
vi /etc/modules
# add drbd
#
Both Servers
lvreduce allows you to reduce the size of a logical volume.
OPTIONS
-L, --size [-]LogicalVolumeSize[bBsSkKmMgGtTpPeE]
With the - sign the value will be subtracted from the logical
volume's actual size.
-r, --resizefs
Resize underlying filesystem together with the logical volume using fsadm(8).
lvreduce --resizefs --size -50G /dev/ubuntu/root
cp /home/user/drbd-8.3.11/scripts/global_common.conf.protoD /etc/drbd.d/global_common.conf
cp /home/user/drbd-8.3.11/scripts/testvms_protoD.res /etc/drbd.d/SystemHA_protoD.res
lvcreate -n test -L 10G ubuntu
vi /etc/drbd.d/SystemHA_protoD.res
#
resource drbd-vm {
device /dev/drbd1;
disk /dev/ubuntu/test;
meta-disk internal;
on cheng-HP-Compaq-Elite-8300-SFF {
address 147.8.177.50:7791;
}
on wang-HP-Compaq-Elite-8300-SFF {
address 147.8.179.243:7791;
}
}
#
#Create the meta-data for the SystemHA-disk and then bring up the resource. Do this on both machines
drbdadm create-md drbd-vm
###answer y or yes for all questions, in the above command
drbdadm up drbd-vm
Server #1 - this will override all of the DRBD data on server #2
drbdadm -- --overwrite-data-of-peer primary drbd-vm
From /proc/drbd you can monitor the current status of the DRBD resource.
Notes:
- When Remus is running you can do 'cat /proc/drbd' and you'll notice the status will be Primary/Primary
- When configuring your VM, use drbd:drbd-vm for the disk. not phy:/… or any other format
- It's one VM per DRBD disk like this, so you have to create a new one for each VM
Both servers
vi /etc/xen/xend-config.sxp
# Ensure the following are uncommented
(xend-relocation-server yes)
(xend-relocation-port 8002)
(xend-relocation-address '')
(xend-relocation-hosts-allow '')
#
vi /etc/modules
#
sch_plug
sch_prio
sch_ingress
cls_basic
cls_tcindex
cls_u32
act_mirred
ifb
#
Server #1
vi /etc/xen/rt.cfg
#
name = "rt"
memory = 256
disk = [ 'drbd:drbd-vm,xvda,w' ]
vif = [ 'mac=18:66:da:03:15:b1' ]
kernel = "/var/lib/xen/images/ubuntu-netboot/vmlinuz"
ramdisk = "/var/lib/xen/images/ubuntu-netboot/initrd.gz"
extra = "debian-installer/exit/always_halt=true -- console=hvc0"
#
xm create /etc/xen/rt.cfg -c
# run the install
vi /etc/xen/rt.cfg
#
bootloader = "/usr/lib/xen/bin/pygrub"
# comment out 'kernel' 'ramdisk' and 'extra'
#
xm create /etc/xen/rt.cfg -c
# see it running
xm list
# reconnect
xm console rt
# immediately terminate
xm destroy rt
Sometimes, for development or analysis purposes, you dont really want to replicate to a physical machine. You just want to gather up the statistics such as number of pages that changed in a checkpoint, size of data sent, etc. In this case, all you need is a system to continuously checkpoint the VM and replicate it to a sink like /dev/null, while still gathering up stats.
#remus options:
-i MS, --interval=MS checkpoint every MS milliseconds
--blackhole replicate to /dev/null (no disk checkpoints, only memory & net buffering)
--no-net run without net buffering (benchmark option)
#
remus -i 40 --blackhole --no-net rt dummyHost >/var/log/xen/domU-blackhole.log 2>&1 &
The VM rt is continuously checkpointed but replicated to /dev/null. Gather up all the stats you want and then kill remus
pkill -USR1 remus
Remus supports most domU guests that can be run in PV, however, you may see this warning when you start Remus:
"WARNING: suspend event channel unavailable, falling back to slow xenstore signalling"
This means the kernel in the guest doesn't have "suspend event channel" support, which in turn basically Remus is going to work, but isn't going to perform well.