![\includegraphics[]{figs/xenlogo.eps}](img1.png)
| Users' Manual |
| Xen v3.0 |
| Users' Manual |
| Xen v3.0 |
DISCLAIMER: This documentation is always under active development and as such there may be mistakes and omissions -- watch out for these and please report any you find to the developers' mailing list, [email protected]. The latest version is always available on-line. Contributions of material, suggestions and corrections are welcome.
Xen is Copyright ©2002-2005, University of Cambridge, UK, XenSource Inc., IBM Corp., Hewlett-Packard Co., Intel Corp., AMD Inc., and others. All rights reserved.
Xen is an open-source project. Most portions of Xen are licensed for copying under the terms of the GNU General Public License, version 2. Other portions are licensed under the terms of the GNU Lesser General Public License, the Zope Public License 2.0, or under ``BSD-style'' licenses. Please refer to the COPYING file for details.
Xen includes software by Christopher Clark. This software is covered by the following licence:
Copyright (c) 2002, Christopher Clark. All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
=10000 =10000 1.1
Xen is an open-source para-virtualizing virtual machine monitor (VMM), or ``hypervisor'', for the x86 processor architecture. Xen can securely execute multiple virtual machines on a single physical system with close-to-native performance. Xen facilitates enterprise-grade functionality, including:
Usage scenarios for Xen include:
Para-virtualization permits very high performance virtualization, even on architectures like x86 that are traditionally very hard to virtualize.
This approach requires operating systems to be ported to run on Xen. Porting an OS to run on Xen is similar to supporting a new hardware platform, however the process is simplified because the para-virtual machine architecture is very similar to the underlying native hardware. Even though operating system kernels must explicitly support Xen, a key feature is that user space applications and libraries do not require modification.
With hardware CPU virtualization as provided by Intel VT and AMD SVM technology, the ability to run an unmodified guest OS kernel is available. No porting of the OS is required, although some additional driver support is necessary within Xen itself. Unlike traditional full virtualization hypervisors, which suffer a tremendous performance overhead, the combination of Xen and VT or Xen and Pacifica technology complement one another to offer superb performance for para-virtualized guest operating systems and full support for unmodified guests running natively on the processor.
Paravirtualized Xen support is available for increasingly many operating systems: currently, mature Linux support is available and included in the standard distribution. Other OS ports--including NetBSD, FreeBSD and Solaris x86 v10--are nearing completion.
Xen currently runs on the x86 architecture, requiring a ``P6'' or newer processor (e.g. Pentium Pro, Celeron, Pentium II, Pentium III, Pentium IV, Xeon, AMD Athlon, AMD Duron). Multiprocessor machines are supported, and there is support for HyperThreading (SMT). In addition, ports to IA64 and Power architectures are supported.
The default 32-bit Xen supports for Intel's Physical Addressing Extensions (PAE), which enable x86/32 machines to address up to 64 GB of physical memory. It also supports non-PAE 32-bit Xen up to 4GB of memory. Xen also supports x86/64 platforms such as Intel EM64T and AMD Opteron which can currently address up to 1TB of physical memory.
Xen offloads most of the hardware support issues to the guest OS running in the Domain 0 management virtual machine. Xen itself contains only the code required to detect and start secondary processors, set up interrupt routing, and perform PCI bus enumeration. Device drivers run within a privileged guest OS rather than within Xen itself. This approach provides compatibility with the majority of device hardware supported by Linux. The default XenLinux build contains support for most server-class network and disk hardware, but you can add support for other hardware by configuring your XenLinux kernel in the normal way.
A Xen system has multiple layers, the lowest and most privileged of which is Xen itself.
Xen may host multiple guest operating systems, each of which is executed within a secure virtual machine. In Xen terminology, a domain. Domains are scheduled by Xen to make effective use of the available physical CPUs. Each guest OS manages its own applications. This management includes the responsibility of scheduling each application within the time allotted to the VM by Xen.
The first domain, domain 0, is created automatically when the system boots and has special management privileges. Domain 0 builds other domains and manages their virtual devices. It also performs administrative tasks such as suspending, resuming and migrating other virtual machines.
Within domain 0, a process called xend runs to manage the system. Xend is responsible for managing virtual machines and providing access to their consoles. Commands are issued to xend over an HTTP interface, via a command-line tool.
Xen was originally developed by the Systems Research Group at the University of Cambridge Computer Laboratory as part of the XenoServers project, funded by the UK-EPSRC.
XenoServers aim to provide a ``public infrastructure for global distributed computing''. Xen plays a key part in that, allowing one to efficiently partition a single machine to enable multiple independent clients to run their operating systems and applications in an environment. This environment provides protection, resource isolation and accounting. The project web page contains further information along with pointers to papers and technical reports: http://www.cl.cam.ac.uk/xeno
Xen has grown into a fully-fledged project in its own right, enabling us to investigate interesting research issues regarding the best techniques for virtualizing resources such as the CPU, memory, disk and network. Project contributors now include XenSource, Intel, IBM, HP, AMD, Novell, RedHat.
Xen was first described in a paper presented at SOSP in 20031.2, and the first public release (1.0) was made that October. Since then, Xen has significantly matured and is now used in production scenarios on many sites.
Xen 3.0.0 offers:
Xen 3.0 features greatly enhanced hardware support, configuration flexibility, usability and a larger complement of supported operating systems. This latest release takes Xen a step closer to being the definitive open source solution for virtualization.
The Xen distribution includes three main components: Xen itself, ports of Linux and NetBSD to run on Xen, and the userspace tools required to manage a Xen-based system. This chapter describes how to install the Xen 3.0 distribution from source. Alternatively, there may be pre-built packages available as part of your operating system distribution.
The following is a full list of prerequisites. Items marked `
' are
required by the xend control tools, and hence required if you want to
run more than one virtual machine; items marked `*' are only required
if you wish to build from source.
Once you have satisfied these prerequisites, you can now install either a binary or source distribution of Xen.
Pre-built tarballs are available for download from the XenSource downloads page:
http://www.xensource.com/downloads/
Once you've downloaded the tarball, simply unpack and install:
# tar zxvf xen-3.0-install.tgz # cd xen-3.0-install # sh ./install.sh
Once you've installed the binaries you need to configure your system as described in Section 2.5.
http://www.xensource.com/downloads/
Once you've downloaded the RPMs, you typically install them via the RPM commands:
# rpm -iv rpmname
See the instructions and the Release Notes for each RPM set referenced at:
http://www.xensource.com/downloads/.
This section describes how to obtain, build and install Xen from source.
The Xen source tree is available as either a compressed source tarball or as a clone of our master Mercurial repository.
http://www.xensource.com/downloads/
http://xenbits.xensource.comSee the instructions and the Getting Started Guide referenced at:
http://www.xensource.com/downloads/
The top-level Xen Makefile includes a target ``world'' that will do the following:
After the build has completed you should have a top-level directory called dist/ in which all resulting targets will be placed. Of particular interest are the two XenLinux kernel images, one with a ``-xen0'' extension which contains hardware device drivers and drivers for Xen's virtual devices, and one with a ``-xenU'' extension that just contains the virtual ones. These are found in dist/install/boot/ along with the image for Xen itself and the configuration files used during the build.
To customize the set of kernels built you need to edit the top-level Makefile. Look for the line:
KERNELS ?= linux-2.6-xen0 linux-2.6-xenU
You can edit this line to include any set of operating system kernels which have configurations in the top-level buildconfigs/ directory.
If you wish to build a customized XenLinux kernel (e.g. to support additional devices or enable distribution-required features), you can use the standard Linux configuration mechanisms, specifying that the architecture being built for is xen, e.g:
# cd linux-2.6.12-xen0 # make ARCH=xen xconfig # cd .. # make
You can also copy an existing Linux configuration (.config) into e.g. linux-2.6.12-xen0 and execute:
# make ARCH=xen oldconfig
You may be prompted with some Xen-specific options. We advise accepting the defaults for these options.
Note that the only difference between the two types of Linux kernels that are built is the configuration file used for each. The ``U'' suffixed (unprivileged) versions don't contain any of the physical hardware device drivers, leading to a 30% reduction in size; hence you may prefer these for your non-privileged domains. The ``0'' suffixed privileged versions can be used to boot the system, as well as in driver domains and unprivileged domains.
The files produced by the build process are stored under the dist/install/ directory. To install them in their default locations, do:
# make install
Alternatively, users with special installation requirements may wish to install them manually by copying the files to their appropriate destinations.
The dist/install/boot directory will also contain the config files used for building the XenLinux kernels, and also versions of Xen and XenLinux kernels that contain debug symbols such as (xen-syms-3.0.0 and vmlinux-syms-2.6.12.6-xen0) which are essential for interpreting crash dumps. Retain these files as the developers may wish to see them if you post on the mailing list.
Once you have built and installed the Xen distribution, it is simple to prepare the machine for booting and running Xen.
An entry should be added to grub.conf (often found under /boot/ or /boot/grub/) to allow Xen / XenLinux to boot. This file is sometimes called menu.lst, depending on your distribution. The entry should look something like the following:
title Xen 3.0 / XenLinux 2.6 kernel /boot/xen-3.0.gz dom0_mem=262144 module /boot/vmlinuz-2.6-xen0 root=/dev/sda4 ro console=tty0
The kernel line tells GRUB where to find Xen itself and what boot parameters should be passed to it (in this case, setting the domain 0 memory allocation in kilobytes and the settings for the serial port). For more details on the various Xen boot parameters see Section 11.3.
The module line of the configuration describes the location of the XenLinux kernel that Xen should start and the parameters that should be passed to it. These are standard Linux parameters, identifying the root device and specifying it be initially mounted read only and instructing that console output be sent to the screen. Some distributions such as SuSE do not require the ro parameter.
To use an initrd, add another module line to the configuration, like:
module /boot/my_initrd.gz
When installing a new kernel, it is recommended that you do not delete existing menu options from menu.lst, as you may wish to boot your old Linux kernel in future, particularly if you have problems.
Serial console access allows you to manage, monitor, and interact with your system over a serial console. This can allow access from another nearby system via a null-modem (``LapLink'') cable or remotely via a serial concentrator.
You system's BIOS, bootloader (GRUB), Xen, Linux, and login access must each be individually configured for serial console access. It is not strictly necessary to have each component fully functional, but it can be quite useful.
For general information on serial console configuration under Linux, refer to the ``Remote Serial Console HOWTO'' at The Linux Documentation Project: http://www.tldp.org
Enabling system serial console output neither enables nor disables serial capabilities in GRUB, Xen, or Linux, but may make remote management of your system more convenient by displaying POST and other boot messages over serial port and allowing remote BIOS configuration.
Refer to your hardware vendor's documentation for capabilities and procedures to enable BIOS serial redirection.
Enabling GRUB serial console output neither enables nor disables Xen or Linux serial capabilities, but may made remote management of your system more convenient by displaying GRUB prompts, menus, and actions over serial port and allowing remote GRUB management.
Adding the following two lines to your GRUB configuration file, typically either /boot/grub/menu.lst or /boot/grub/grub.conf depending on your distro, will enable GRUB serial output.
serial --unit=0 --speed=115200 --word=8 --parity=no --stop=1 terminal --timeout=10 serial console
Note that when both the serial port and the local monitor and keyboard are enabled, the text ``Press any key to continue'' will appear at both. Pressing a key on one device will cause GRUB to display to that device. The other device will see no output. If no key is pressed before the timeout period expires, the system will boot to the default GRUB boot entry.
Please refer to the GRUB documentation for further information.
Enabling Xen serial console output neither enables nor disables Linux kernel output or logging in to Linux over serial port. It does however allow you to monitor and log the Xen boot process via serial console and can be very useful in debugging.
In order to configure Xen serial console output, it is necessary to add a boot option to your GRUB config; e.g. replace the previous example kernel line with:
kernel /boot/xen.gz dom0_mem=131072 com1=115200,8n1 console=com1,vga
This configures Xen to output on COM1 at 115,200 baud, 8 data bits, no parity and 1 stop bit. Modify these parameters for your environment. See Section 11.3 for an explanation of all boot parameters.
One can also configure XenLinux to share the serial console; to achieve this append ``console=ttyS0'' to your module line.
Enabling Linux serial console output at boot neither enables nor disables logging in to Linux over serial port. It does however allow you to monitor and log the Linux boot process via serial console and can be very useful in debugging.
To enable Linux output at boot time, add the parameter console=ttyS0 (or ttyS1, ttyS2, etc.) to your kernel GRUB line. Under Xen, this might be:
module /vmlinuz-2.6-xen0 ro root=/dev/VolGroup00/LogVol00 \ console=ttyS0, 115200
to enable output over ttyS0 at 115200 baud.
Logging in to Linux via serial console, under Xen or otherwise, requires specifying a login prompt be started on the serial port. To permit root logins over serial console, the serial port must be added to /etc/securetty.
To automatically start a login prompt over the serial port, add the line:
c:2345:respawn:/sbin/mingetty ttyS0to /etc/inittab. Run init q to force a reload of your inttab and start getty.
To enable root logins, add ttyS0 to /etc/securetty if not already present.
Your distribution may use an alternate getty; options include getty, mgetty and agetty. Consult your distribution's documentation for further information.
Users of the XenLinux 2.6 kernel should disable Thread Local Storage (TLS) (e.g. by doing a mv /lib/tls /lib/tls.disabled) before attempting to boot a XenLinux kernel2.4. You can always reenable TLS by restoring the directory to its original location (i.e. mv /lib/tls.disabled /lib/tls).
The reason for this is that the current TLS implementation uses segmentation in a way that is not permissible under Xen. If TLS is not disabled, an emulation mode is used within Xen which reduces performance substantially. To ensure full performance you should install a `Xen-friendly' (nosegneg) version of the library.
It should now be possible to restart the system and use Xen. Reboot and choose the new Xen option when the Grub screen appears.
What follows should look much like a conventional Linux boot. The first portion of the output comes from Xen itself, supplying low level information about itself and the underlying hardware. The last portion of the output comes from XenLinux.
You may see some error messages during the XenLinux boot. These are not necessarily anything to worry about--they may result from kernel configuration differences between your XenLinux kernel and the one you usually use.
When the boot completes, you should be able to log into your system as usual. If you are unable to log in, you should still be able to reboot with your normal Linux kernel by selecting it at the GRUB prompt.
Booting the system into Xen will bring you up into the privileged management domain, Domain0. At that point you are ready to create guest domains and ``boot'' them using the xm create command.
After installation and configuration is complete, reboot the system and and choose the new Xen option when the Grub screen appears.
What follows should look much like a conventional Linux boot. The first portion of the output comes from Xen itself, supplying low level information about itself and the underlying hardware. The last portion of the output comes from XenLinux.
When the boot completes, you should be able to log into your system as usual. If you are unable to log in, you should still be able to reboot with your normal Linux kernel by selecting it at the GRUB prompt.
The first step in creating a new domain is to prepare a root filesystem for it to boot. Typically, this might be stored in a normal partition, an LVM or other volume manager partition, a disk file or on an NFS server. A simple way to do this is simply to boot from your standard OS install CD and install the distribution into another partition on your hard drive.
To start the xend control daemon, type
# xend start
If you wish the daemon to start automatically, see the instructions in Section 4.1. Once the daemon is running, you can use the xm tool to monitor and maintain the domains running on your system. This chapter provides only a brief tutorial. We provide full details of the xm tool in the next chapter.
Before you can start an additional domain, you must create a configuration file. We provide two example files which you can use as a starting point:
There are also a number of other examples which you may find useful. Copy one of these files and edit it as appropriate. Typical values you may wish to edit include:
You may also want to edit the vif variable in order to choose the MAC address of the virtual ethernet interface yourself. For example:
vif = ['mac=00:16:3E:F6:BB:B3']
If you do not set this variable, xend will automatically generate a
random MAC address from the range 00:16:3E:xx:xx:xx, assigned by IEEE to
XenSource as an OUI (organizationally unique identifier). XenSource
Inc. gives permission for anyone to use addresses randomly allocated
from this range for use by their Xen domains.
For a list of IEEE OUI assignments, see http://standards.ieee.org/regauth/oui/oui.txt
The xm tool provides a variety of commands for managing domains. Use the create command to start new domains. Assuming you've created a configuration file myvmconf based around /etc/xen/xmexample2, to start a domain with virtual machine ID 1 you should type:
# xm create -c myvmconf vmid=1
The -c switch causes xm to turn into the domain's console after creation. The vmid=1 sets the vmid variable used in the myvmconf file.
You should see the console boot messages from the new domain appearing in the terminal in which you typed the command, culminating in a login prompt.
It is possible to have certain domains start automatically at boot time and to have dom0 wait for all running domains to shutdown before it shuts down the system.
To specify a domain is to start at boot-time, place its configuration file (or a link to it) under /etc/xen/auto/.
A Sys-V style init script for Red Hat and LSB-compliant systems is provided and will be automatically copied to /etc/init.d/ during install. You can then enable it in the appropriate way for your distribution.
For instance, on Red Hat:
# chkconfig --add xendomains
By default, this will start the boot-time domains in runlevels 3, 4 and 5.
You can also use the service command to run this script manually, e.g:
# service xendomains start
Starts all the domains with config files under /etc/xen/auto/.
# service xendomains stop
Shuts down all running Xen domains.
This chapter summarizes the management software and tools available.
The Xend node control daemon performs system management functions related to virtual machines. It forms a central point of control of virtualized resources, and must be running in order to start and manage virtual machines. Xend must be run as root because it needs access to privileged system management functions.
An initialization script named /etc/init.d/xend is provided to start Xend at boot time. Use the tool appropriate (i.e. chkconfig) for your Linux distribution to specify the runlevels at which this script should be executed, or manually create symbolic links in the correct runlevel directories.
Xend can be started on the command line as well, and supports the following set of parameters:
# xend start |
start xend, if not already running |
# xend stop |
stop xend if already running |
# xend restart |
restart xend if running, otherwise start it |
# xend status |
indicates xend status by its return code |
A SysV init script called xend is provided to start xend at boot time. make install installs this script in /etc/init.d. To enable it, you have to make symbolic links in the appropriate runlevel directories or use the chkconfig tool, where available. Once xend is running, administration can be done using the xm tool.
As xend runs, events will be logged to /var/log/xen/xend.log and (less frequently) to /var/log/xen/xend-debug.log. These, along with the standard syslog files, are useful when troubleshooting problems.
Xend is written in Python. At startup, it reads its configuration information from the file /etc/xen/xend-config.sxp. The Xen installation places an example xend-config.sxp file in the /etc/xen subdirectory which should work for most installations.
See the example configuration file xend-debug.sxp and the section 5 man page xend-config.sxp for a full list of parameters and more detailed information. Some of the most important parameters are discussed below.
An HTTP interface and a Unix domain socket API are available to communicate with Xend. This allows remote users to pass commands to the daemon. By default, Xend does not start an HTTP server. It does start a Unix domain socket management server, as the low level utility xm requires it. For support of cross-machine migration, Xend can start a relocation server. This support is not enabled by default for security reasons.
Note: the example xend configuration file modifies the defaults and starts up Xend as an HTTP server as well as a relocation server.
From the file:
#(xend-http-server no) (xend-http-server yes) #(xend-unix-server yes) #(xend-relocation-server no) (xend-relocation-server yes)
Comment or uncomment lines in that file to disable or enable features that you require.
Connections from remote hosts are disabled by default:
# Address xend should listen on for HTTP connections, if xend-http-server is # set. # Specifying 'localhost' prevents remote connections. # Specifying the empty string '' (the default) allows all connections. #(xend-address '') (xend-address localhost)
It is recommended that if migration support is not needed, the xend-relocation-server parameter value be changed to ``no'' or commented out.
The xm tool is the primary tool for managing Xen from the console. The general format of an xm command line is:
# xm command [switches] [arguments] [variables]
The available switches and arguments are dependent on the command chosen. The variables may be set using declarations of the form variable=value and command line declarations override any of the values in the configuration file being used, including the standard variables described above and any custom variables (for instance, the xmdefconfig file uses a vmid variable).
For online help for the commands available, type:
# xm help
This will list the most commonly used commands. The full list can be obtained
using xm help --long. You can also type xm help <command>
for more information on a given command.
One useful command is # xm list which lists all domains running in rows
of the following format:
The meaning of each field is as follows:
The xm list command also supports a long output format when the -l switch is used. This outputs the full details of the running domains in xend's SXP configuration format.
If you want to know how long your domains have been running for, then
you can use the # xm uptime command.
You can get access to the console of a particular domain using
the # xm console command (e.g. # xm console myVM).
The credit CPU scheduler automatically load balances guest VCPUs across all available physical CPUs on an SMP host. The user need not manually pin VCPUs to load balance the system. However, she can restrict which CPUs a particular VCPU may run on using the xm vcpu-pin command.
Each guest domain is assigned a weight and a cap.
A domain with a weight of 512 will get twice as much CPU as a domain with a weight of 256 on a contended host. Legal weights range from 1 to 65535 and the default is 256.
The cap optionally fixes the maximum amount of CPU a guest will be able to consume, even if the host system has idle CPU cycles. The cap is expressed in percentage of one physical CPU: 100 is 1 physical CPU, 50 is half a CPU, 400 is 4 CPUs, etc... The default, 0, means there is no upper cap.
When you are running with the credit scheduler, you can check and modify your domains' weights and caps using the xm sched-credit command:
xm sched-credit -d <domain> |
lists weight and cap |
xm sched-credit -d <domain> -w <weight> |
sets the weight |
xm sched-credit -d <domain> -c <cap> |
sets the cap |
The following contains the syntax of the domain configuration files and description of how to further specify networking, driver domain and general scheduling behavior.
Xen configuration files contain the following standard variables. Unless otherwise stated, configuration items should be enclosed in quotes: see the configuration scripts in /etc/xen/ for concrete examples.
vif = [ 'mac=00:16:3E:00:00:11, bridge=xen-br0',
'bridge=xen-br1' ]
to assign a MAC address and bridge to the first interface and assign
a different bridge to the second interface, leaving xend to choose
the MAC address. The settings that may be overridden in this way are
type, mac, bridge, ip, script, backend, and vifname.
disk = [ 'phy:hda1,sda1,r' ]
exports physical device /dev/hda1 to the domain as
/dev/sda1 with read-only access. Exporting a disk read-write
which is currently mounted is dangerous - if you are certain
you wish to do this, you can specify w! as the mode.
Additional fields are documented in the example configuration files (e.g. to configure virtual TPM functionality).
For additional flexibility, it is also possible to include Python scripting commands in configuration files. An example of this is the xmexample2 file, which uses Python code to handle the vmid variable.
For many users, the default installation should work ``out of the box''. More complicated network setups, for instance with multiple Ethernet interfaces and/or existing bridging setups will require some special configuration.
The purpose of this section is to describe the mechanisms provided by xend to allow a flexible configuration for Xen's virtual networking.
Each domain network interface is connected to a virtual network interface in dom0 by a point to point link (effectively a ``virtual crossover cable''). These devices are named vif<domid>.<vifid> (e.g. vif1.0 for the first interface in domain 1, vif3.1 for the second interface in domain 3).
Traffic on these virtual interfaces is handled in domain 0 using standard Linux mechanisms for bridging, routing, rate limiting, etc. Xend calls on two shell scripts to perform initial configuration of the network and configuration of new virtual interfaces. By default, these scripts configure a single bridge for all the virtual interfaces. Arbitrary routing / bridging configurations can be configured by customizing the scripts, as described in the following section.
Xen's virtual networking is configured by two shell scripts (by default network-bridge and vif-bridge). These are called automatically by xend when certain events occur, with arguments to the scripts providing further contextual information. These scripts are found by default in /etc/xen/scripts. The names and locations of the scripts can be configured in /etc/xen/xend-config.sxp.
Other example scripts are available (network-route and vif-route, network-nat and vif-nat). For more complex network setups (e.g. where routing is required or integrate with existing bridges) these scripts may be replaced with customized variants for your site's preferred configuration.
Individual PCI devices can be assigned to a given domain (a PCI driver domain) to allow that domain direct access to the PCI hardware.
While PCI Driver Domains can increase the stability and security of a system by addressing a number of security concerns, there are some security issues that remain that you can read about in Section 9.2.
An example kernel command-line which hides two PCI devices might be:
Some examples:
Bind a device to the PCI Backend which is not bound to any other driver.
# # Add a new slot to the PCI Backend's list # echo -n 0000:01:04.d > /sys/bus/pci/drivers/pciback/new_slot # # Now that the backend is watching for the slot, bind to it # echo -n 0000:01:04.d > /sys/bus/pci/drivers/pciback/bind
Unbind a device from its driver and bind to the PCI Backend.
# # Unbind a PCI network card from its network driver # echo -n 0000:05:02.0 > /sys/bus/pci/drivers/3c905/unbind # # And now bind it to the PCI Backend # echo -n 0000:05:02.0 > /sys/bus/pci/drivers/pciback/new_slot # echo -n 0000:05:02.0 > /sys/bus/pci/drivers/pciback/bind
Note that the "-n" option in the example is important as it causes echo to not output a new-line.
The policy file is heavily commented and is intended to provide enough documentation for developers to extend it.
Currently, the only way to reset the permissive flag is to unbind the device from the PCI Backend driver.
You may notice that every device bound to the PCI backend has 17 quirks standard "quirks" regardless of xend-pci-quirks.sxp. These default entries are necessary to support interactions between the PCI bus manager and the device bound to it. Even non-quirky devices should have these standard entries.
In this case, preference was given to accuracy over aesthetics by choosing to show the standard quirks in the quirks list rather than hide them from the inquiring user
(device (pci
(dev (domain 0x0)(bus 0x3)(slot 0x1a)(func 0x1)
(dev (domain 0x0)(bus 0x1)(slot 0x5)(func 0x0)
)
Paravirtualized domains can be given access to a virtualized version of a TPM. This enables applications in these domains to use the services of the TPM device for example through a TSS stack 5.1. The Xen source repository provides the necessary software components to enable virtual TPM access. Support is provided through several different pieces. First, a TPM emulator has been modified to provide TPM's functionality for the virtual TPM subsystem. Second, a virtual TPM Manager coordinates the virtual TPMs efforts, manages their creation, and provides protected key storage using the TPM. Third, a device driver pair providing a TPM front- and backend is available for XenLinux to deliver TPM commands from the domain to the virtual TPM manager, which dispatches it to a software TPM. Since the TPM Manager relies on a HW TPM for protected key storage, therefore this subsystem requires a Linux-supported hardware TPM. For development purposes, a TPM emulator is available for use on non-TPM enabled platforms.
modprobe tpmbk
Similarly, the TPM frontend driver must be compiled for the kernel trying to use TPM functionality. Its driver can be selected in the kernel configuration section Device Driver / Character Devices / TPM Devices. Along with that the TPM driver for the built-in TPM must be selected. If the virtual TPM driver has been compiled as module, it must be activated using the following command:
modprobe tpm_xenu
Furthermore, it is necessary to build the virtual TPM manager and software TPM by making changes to entries in Xen build configuration files. The following entry in the file Config.mk in the Xen root source directory must be made:
VTPM_TOOLS ?= y
After a build of the Xen tree and a reboot of the machine, the TPM backend drive must be loaded. Once loaded, the virtual TPM manager daemon must be started before TPM-enabled guest domains may be launched. To enable being the destination of a virtual TPM Migration, the virtual TPM migration daemon must also be loaded.
vtpm_managerd
vtpm_migratord
Once the VTPM manager is running, the VTPM can be accessed by loading the front end driver in a guest domain.
BUILD_EMULATOR = y
Second, the entry in the file tool/vtpm_manager/Rules.mk must be uncommented as follows:
# TCS talks to fifo's rather than /dev/tpm. TPM Emulator assumed on fifos CFLAGS += -DDUMMY_TPM
Before starting the virtual TPM Manager, start the emulator by executing the following in dom0:
tpm_emulator clear
vtpm = ['instance=<instance number>, backend=<domain id>']
The instance number reflects the preferred virtual TPM instance to associate with the domain. If the selected instance is already associated with another domain, the system will automatically select the next available instance. An instance number greater than zero must be provided. It is possible to omit the instance parameter from the configuration file.
The domain id provides the ID of the domain where the virtual TPM backend driver and virtual TPM are running in. It should currently always be set to '0'.
Examples for valid vtpm entries in the configuration file are
vtpm = ['instance=1, backend=0']and
vtpm = ['backend=0'].
Access to TPM functionality is provided by the virtual TPM frontend driver. Similar to existing hardware TPM drivers, this driver provides basic TPM status information through the sysfs filesystem. In a Xen user domain the sysfs entries can be found in /sys/devices/xen/vtpm-0.
Commands can be sent to the virtual TPM instance using the character device /dev/tpm0 (major 10, minor 224).
Storage can be made available to virtual machines in a number of different ways. This chapter covers some possible configurations.
The most straightforward method is to export a physical block device (a hard drive or partition) from dom0 directly to the guest domain as a virtual block device (VBD).
Storage may also be exported from a filesystem image or a partitioned filesystem image as a file-backed VBD.
Finally, standard network storage protocols such as NBD, iSCSI, NFS, etc., can be used to provide storage to virtual machines.
One of the simplest configurations is to directly export individual partitions from domain 0 to other domains. To achieve this use the phy: specifier in your domain configuration file. For example a line like
disk = ['phy:hda3,sda1,w']
specifies that the partition /dev/hda3 in domain 0 should be
exported read-write to the new domain as /dev/sda1; one could
equally well export it as /dev/hda or /dev/sdb5 should
one wish.
In addition to local disks and partitions, it is possible to export any device that Linux considers to be ``a disk'' in the same manner. For example, if you have iSCSI disks or GNBD volumes imported into domain 0 you can export these to other domains using the phy: disk syntax. E.g.:
disk = ['phy:vg/lvm1,sda2,w']
Block devices should typically only be shared between domains in a read-only fashion otherwise the Linux kernel's file systems will get very confused as the file system structure may change underneath them (having the same ext3 partition mounted rw twice is a sure fire way to cause irreparable damage)! Xend will attempt to prevent you from doing this by checking that the device is not mounted read-write in domain 0, and hasn't already been exported read-write to another domain. If you want read-write sharing, export the directory to other domains via NFS from domain 0 (or use a cluster file system such as GFS or ocfs2).
It is also possible to use a file in Domain 0 as the primary storage for a virtual machine. As well as being convenient, this also has the advantage that the virtual block device will be sparse -- space will only really be allocated as parts of the file are used. So if a virtual machine uses only half of its disk space then the file really takes up half of the size allocated.
For example, to create a 2GB sparse file-backed virtual block device (actually only consumes no disk space at all):
# dd if=/dev/zero of=vm1disk bs=1k seek=2048k count=0
Make a file system in the disk file:
# mkfs -t ext3 vm1disk
(when the tool asks for confirmation, answer `y')
Populate the file system e.g. by copying from the current root:
# mount -o loop vm1disk /mnt
# cp -ax /{root,dev,var,etc,usr,bin,sbin,lib} /mnt
# mkdir /mnt/{proc,sys,home,tmp}
Tailor the file system by editing /etc/fstab, /etc/hostname, etc. Don't forget to edit the files in the mounted file system, instead of your domain 0 filesystem, e.g. you would edit /mnt/etc/fstab instead of /etc/fstab. For this example put /dev/sda1 to root in fstab.
Now unmount (this is important!):
# umount /mnt
In the configuration file set:
disk = ['tap:aio:/full/path/to/vm1disk,sda1,w']
As the virtual machine writes to its `disk', the sparse file will be filled in and consume more space up to the original 2GB.
Note: Users that have worked with file-backed VBDs on Xen in previous versions will be interested to know that this support is now provided through the blktap driver instead of the loopback driver. This change results in file-based block devices that are higher-performance, more scalable, and which provide better safety properties for VBD data. All that is required to update your existing file-backed VM configurations is to change VBD configuration lines from:
disk = ['file:/full/path/to/vm1disk,sda1,w']
to:
disk = ['tap:aio:/full/path/to/vm1disk,sda1,w']
Note: Loopback mounted VBDs have now been replaced with blktap-based support for raw image files, as described above. This section remains to detail a configuration that was used by older Xen versions.
Raw image file-backed VBDs may also be attached to VMs using the Linux loopback driver. The only required change to the raw file instructions above are to specify the configuration entry as:
disk = ['file:/full/path/to/vm1disk,sda1,w']
Note that loopback file-backed VBDs may not be appropriate for backing I/O-intensive domains. This approach is known to experience substantial slowdowns under heavy I/O workloads, due to the I/O handling by the loopback block device used to support file-backed VBDs in dom0. Loopback support remains for old Xen installations, and users are strongly encouraged to use the blktap-based file support (using ``tap:aio'' as described above).
Additionally, Linux supports a maximum of eight loopback file-backed VBDs across all domains by default. This limit can be statically increased by using the max_loop module parameter if CONFIG_BLK_DEV_LOOP is compiled as a module in the dom0 kernel, or by using the max_loop=n boot option if CONFIG_BLK_DEV_LOOP is compiled directly into the dom0 kernel. Again, users are encouraged to use the blktap-based file support described above which scales to much larger number of active VBDs.
A particularly appealing solution is to use LVM volumes as backing for domain file-systems since this allows dynamic growing/shrinking of volumes as well as snapshot and other features.
To initialize a partition to support LVM volumes:
# pvcreate /dev/sda10
Create a volume group named `vg' on the physical partition:
# vgcreate vg /dev/sda10
Create a logical volume of size 4GB named `myvmdisk1':
# lvcreate -L4096M -n myvmdisk1 vg
You should now see that you have a /dev/vg/myvmdisk1 Make a filesystem, mount it and populate it, e.g.:
# mkfs -t ext3 /dev/vg/myvmdisk1 # mount /dev/vg/myvmdisk1 /mnt # cp -ax / /mnt # umount /mnt
Now configure your VM with the following disk configuration:
disk = [ 'phy:vg/myvmdisk1,sda1,w' ]
LVM enables you to grow the size of logical volumes, but you'll need to resize the corresponding file system to make use of the new space. Some file systems (e.g. ext3) now support online resize. See the LVM manuals for more details.
You can also use LVM for creating copy-on-write (CoW) clones of LVM volumes (known as writable persistent snapshots in LVM terminology). This facility is new in Linux 2.6.8, so isn't as stable as one might hope. In particular, using lots of CoW LVM disks consumes a lot of dom0 memory, and error conditions such as running out of disk space are not handled well. Hopefully this will improve in future.
To create two copy-on-write clones of the above file system you would use the following commands:
# lvcreate -s -L1024M -n myclonedisk1 /dev/vg/myvmdisk1 # lvcreate -s -L1024M -n myclonedisk2 /dev/vg/myvmdisk1
Each of these can grow to have 1GB of differences from the master volume. You can grow the amount of space for storing the differences using the lvextend command, e.g.:
# lvextend +100M /dev/vg/myclonedisk1
Don't let the `differences volume' ever fill up otherwise LVM gets rather confused. It may be possible to automate the growing process by using dmsetup wait to spot the volume getting full and then issue an lvextend.
In principle, it is possible to continue writing to the volume that has been cloned (the changes will not be visible to the clones), but we wouldn't recommend this: have the cloned volume as a `pristine' file system install that isn't mounted directly by any of the virtual machines.
First, populate a root filesystem in a directory on the server machine. This can be on a distinct physical machine, or simply run within a virtual machine on the same node.
Now configure the NFS server to export this filesystem over the network by adding a line to /etc/exports, for instance:
/export/vm1root 192.0.2.4/24 (rw,sync,no_root_squash)
Finally, configure the domain to use NFS root. In addition to the normal variables, you should make sure to set the following values in the domain's configuration file:
root = '/dev/nfs' nfs_server = '2.3.4.5' # substitute IP address of server nfs_root = '/path/to/root' # path to root FS on the server
The domain will need network access at boot time, so either statically configure an IP address using the config variables ip, netmask, gateway, hostname; or enable DHCP (dhcp='dhcp').
Note that the Linux NFS root implementation is known to have stability problems under high load (this is not a Xen-specific problem), so this configuration may not be appropriate for critical servers.
Xen allows a domain's virtual CPU(s) to be associated with one or more host CPUs. This can be used to allocate real resources among one or more guests, or to make optimal use of processor resources when utilizing dual-core, hyperthreading, or other advanced CPU technologies.
Xen enumerates physical CPUs in a `depth first' fashion. For a system with both hyperthreading and multiple cores, this would be all the hyperthreads on a given core, then all the cores on a given socket, and then all sockets. I.e. if you had a two socket, dual core, hyperthreaded Xeon the CPU order would be:
| socket0 | socket1 | ||||||
| core0 | core1 | core0 | core1 | ||||
| ht0 | ht1 | ht0 | ht1 | ht0 | ht1 | ht0 | ht1 |
| #0 | #1 | #2 | #3 | #4 | #5 | #6 | #7 |
Having multiple vcpus belonging to the same domain mapped to the same physical CPU is very likely to lead to poor performance. It's better to use `vcpus-set' to hot-unplug one of the vcpus and ensure the others are pinned on different CPUs.
If you are running IO intensive tasks, its typically better to dedicate either a hyperthread or whole core to running domain 0, and hence pin other domains so that they can't use CPU 0. If your workload is mostly compute intensive, you may want to pin vcpus such that all physical CPU threads are available for guest domains.
The administrator of a Xen system may suspend a virtual machine's current state into a disk file in domain 0, allowing it to be resumed at a later time.
For example you can suspend a domain called ``VM1'' to disk using the command:
# xm save VM1 VM1.chk
This will stop the domain named ``VM1'' and save its current state into a file called VM1.chk.
To resume execution of this domain, use the xm restore command:
# xm restore VM1.chk
This will restore the state of the domain and resume its execution. The domain will carry on as before and the console may be reconnected using the xm console command, as described earlier.
Migration is used to transfer a domain between physical hosts. There are two varieties: regular and live migration. The former moves a virtual machine from one host to another by pausing it, copying its memory contents, and then resuming it on the destination. The latter performs the same logical functionality but without needing to pause the domain for the duration. In general when performing live migration the domain continues its usual activities and--from the user's perspective--the migration should be imperceptible.
To perform a live migration, both hosts must be running Xen / xend and the destination host must have sufficient resources (e.g. memory capacity) to accommodate the domain after the move. Furthermore we currently require both source and destination machines to be on the same L2 subnet.
Currently, there is no support for providing automatic remote access to filesystems stored on local disk when a domain is migrated. Administrators should choose an appropriate storage solution (i.e. SAN, NAS, etc.) to ensure that domain filesystems are also available on their destination node. GNBD is a good method for exporting a volume from one machine to another. iSCSI can do a similar job, but is more complex to set up.
When a domain migrates, it's MAC and IP address move with it, thus it is only possible to migrate VMs within the same layer-2 network and IP subnet. If the destination node is on a different subnet, the administrator would need to manually configure a suitable etherip or IP tunnel in the domain 0 of the remote node.
A domain may be migrated using the xm migrate command. To live migrate a domain to another machine, we would use the command:
# xm migrate --live mydomain destination.ournetwork.com
Without the -live flag, xend simply stops the domain and copies the memory image over to the new node and restarts it. Since domains can have large allocations this can be quite time consuming, even on a Gigabit network. With the -live flag xend attempts to keep the domain running while the migration is in progress, resulting in typical down times of just 60-300ms.
For now it will be necessary to reconnect to the domain's console on the new machine using the xm console command. If a migrated domain has any open network connections then they will be preserved, so SSH connections do not have this limitation.
This chapter describes how to secure a Xen system. It describes a number of scenarios and provides a corresponding set of best practices. It begins with a section devoted to understanding the security implications of a Xen system.
When deploying a Xen system, one must be sure to secure the management domain (Domain-0) as much as possible. If the management domain is compromised, all other domains are also vulnerable. The following are a set of best practices for Domain-0:
Driver domains address a range of security problems that exist regarding the use of device drivers and hardware. On many operating systems in common use today, device drivers run within the kernel with the same privileges as the kernel. Few or no mechanisms exist to protect the integrity of the kernel from a misbehaving (read "buggy") or malicious device driver. Driver domains exist to aid in isolating a device driver within its own virtual machine where it cannot affect the stability and integrity of other domains. If a driver crashes, the driver domain can be restarted rather than have the entire machine crash (and restart) with it. Drivers written by unknown or untrusted third-parties can be confined to an isolated space. Driver domains thus address a number of security and stability issues with device drivers.
However, due to limitations in current hardware, a number of security concerns remain that need to be considered when setting up driver domains (it should be noted that the following list is not intended to be exhaustive).
In this scenario, each node has two network cards in the cluster. One network card is connected to the outside world and one network card is a physically isolated management network specifically for Xen instances to use.
As long as all of the management partitions are trusted equally, this is the most secure scenario. No additional configuration is needed other than forcing Xend to bind to the management interface for relocation.
In this scenario, each node has only one network card but the entire cluster sits behind a firewall. This firewall should do at least the following:
The following iptables rules can be used on each node to prevent migrations to that node from outside the subnet assuming the main firewall does not do this for you:
# this command disables all access to the Xen relocation
# port:
iptables -A INPUT -p tcp --destination-port 8002 -j REJECT
# this command enables Xen relocations only from the specific
# subnet:
iptables -I INPUT -p tcp -{}-source 192.0.2.0/24 \
--destination-port 8002 -j ACCEPT
Migration on an untrusted subnet is not safe in current versions of Xen. It may be possible to perform migrations through a secure tunnel via an VPN or SSH. The only safe option in the absence of a secure tunnel is to disable migration completely. The easiest way to do this is with iptables:
# this command disables all access to the Xen relocation port
iptables -A INPUT -p tcp -{}-destination-port 8002 -j REJECT
Benefits of configuring sHype/ACM in Xen include:
These benefits are very valuable because today's operating systems become increasingly complex and often have no or insufficient mandatory access controls. (Discretionary access controls, supported by most operating systems, are not effective against viruses or misbehaving programs.) Where mandatory access control exists (e.g., SELinux), they usually deploy platform-specific, complex, and difficult to understand security policies. Multi-tier applications in business environments typically require different operating systems (e.g., AIX, Windows, Linux) in different tiers. Related distributed transactions and workloads cannot be easily protected on the OS level. The Xen access control framework steps in to offer a coarse-grained but very robust and consistent security layer and safety net across different platforms and operating systems.
To control sharing between domains, Xen mediates all inter-domain communication (shared memory, events) as well as the access of domains to resources such as storage disks. Thus, Xen can confine distributed workloads (domain payloads) by permitting sharing among domains running the same type of workload and denying sharing between pairs of domains that run different workload types. We assume that-from a Xen perspective-only one workload type is running per user domain. To enable Xen to associate domains and resources with workload types, security labels including the workload types are attached to domains and resources. These labels and the hypervisor sHype controls cannot be manipulated or bypassed by user domains and are effective even against compromised or rogue domains.
![\includegraphics[width=13cm]{figs/acm_overview.eps}](img4.png)
|
First, the sHype/ACM access control must be enabled in the Xen distribution and the distribution must be built and installed (cf Subsection 10.2.1). Before we can enforce security, a Xen security policy must be created (cf Subsection 10.2.2) and deployed (cf Subsection 10.2.3). This policy defines the workload types differentiated during access control. It also defines the rules that compare workload types of domains and resources to decide about access requests. Workload types are represented by security labels that can be securely associated to domains and resources (cf Subsections 10.2.4 and 10.2.5). The functioning of the active sHype/Xen workload protection is demonstrated using simple resource assignment, and domain creation tests in Subsection 10.2.6. Section 10.3 describes the syntax and semantics of the sHype/Xen security policy in detail and introduces briefly the tools that are available to help you create your own sHype security policies.
The next section describes all the necessary steps to create, deploy, and test a simple workload protection policy. It is meant to enable Xen users and developers to quickly try out the sHype/Xen workload protection. Those readers who are interested in learning more about how the sHype access control in Xen works and how it is configured using the XML security policy should read Section 10.3 as well. Section 10.4 concludes this chapter with current limitations of the sHype implementation for Xen.
You are about to configure and deploy the Xen sHype workload protection by following 5 simple steps:
To enable sHype/ACM in Xen, please edit the Config.mk file in the top Xen directory.
(1) In Config.mk
Change: XSM_ENABLE ?= n
To: XSM_ENABLE ?= y
Change: ACM_SECURITY ?= n
To: ACM_SECURITY ?= y
Then install the security-enabled Xen environment as follows:
(2) # make world
# make install
Reboot into the security-enabled Xen hypervisor.
(3) # reboot
Xen will boot into the default security policy. After reboot, you can explore the simple DEFAULT policy.
# xm getpolicy Supported security subsystems : ACM Policy name : DEFAULT Policy type : ACM Version of XML policy : 1.0 Policy configuration : loaded # xm labels SystemManagement # xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 941 1 r----- 38.1 ACM:DEFAULT:SystemManagement
In this state, no domains can be started. Now, a policy can be created and loaded into the hypervisor.
We will use the ezPolicy tool to quickly create a policy that protects
workloads. You will need both the Python and wxPython packages to run
this tool. To run the tool in Domain-0, you can download the wxPython
package from www.wxpython.org or use the command yum install wxPython
in Redhat/Fedora. To run the tool on MS Windows, you also need to download
the Python package from www.python.org. After these packages are installed,
start the ezPolicy tool with the following command:
(4) # xensec_ezpolicy
Figure 10.2 shows a screen-shot of the tool. The
following steps illustrate how you can create the workload definition
shown in Figure 10.2. You can use <CTRL>-h to
pop up a help window at any time. The indicators (a), (b), and (c) in
Figure 10.2 show the buttons that are used during the
3 steps of creating a policy:
To ease the transition from an unlabeled to a fully labeled workload-protection environment, we have added support to sHype/Xen to run unlabeled domains accessing unlabeled resources in addition to labeled domains accessing labeled resources.
Support for running unlabeled domains on sHype/Xen is enabled by adding the
predefined workload type and label __UNLABELED__ to the security
policy. (This is a double underscore
followed by the string ''UNLABELED'' followed by a double underscore.)
The ezPolicy tool automatically adds this organization-level workload type
to a new workload definition (cf Figure 10.2). It can simply be
deleted from the workload definition if no such support is desired. If unlabeled domains
are supported in the policy, then any domain or resource that has no label will implicitly
inherit this label when access control decisions are made. In effect, unlabeled
domains and resources define a new workload type __UNLABELED__, which is
confined from any other labeled workload.
Please use now the ``New Org'' button to add the organization workload types ``A-Bank'', ``B-Bank'', and ``AutoCorp''.
You can refine an organization to differentiate between multiple
department workloads by right-clicking the organization and selecting
Add Department (or selecting an organization and pressing
<CRTL>-a). Create department workloads ``SecurityUnderwriting'',
and ``MarketAnalysis'' for the ``A-Bank''. The resulting layout of the
tool should be similar to the left panel shown in
Figure 10.2.
![\includegraphics[width=13cm]{figs/acm_ezpolicy_gui.eps}](img5.png)
__UNLABELED__
workload type in all run-time exclusion rules because any workload
could run inside unlabeled domains.
To prevent A-Bank and B-Bank workloads (including their
departmental workloads) from running simultaneously on the same
hypervisor system, select the organization ``A-Bank'' and, while
pressing the <CTRL>-key, select the organization ``B-Bank''.
Being cautious, we also prevent unlabeled workloads from running with
any of those workloads by pressing the <CTRL>-key and selecting
``__UNLABELED__''. Now press the button named ``Create run-time exclusion
rule from selection''. A popup window will ask for the name for this run-time
exclusion rule (enter a name or just hit <ENTER>). A rule will
appear on the right panel. The name is used as reference only and does
not affect access control decisions.
Please repeat this process to create another run-time exclusion rule for the department workloads ``A-Bank.SecurityUnderwriting'', ``A-Bank.MarketAnalysis''. Also add the ``__UNLABELED__'' workload type to this conflict set.
The resulting layout of your window should be similar to Figure 10.2. Save this workload definition by selecting ``Save Workload Definition as ...'' in the ``File'' menu. This workload definition can be later refined if required.
mytest. If you are running ezPolicy in
Domain-0, the resulting policy file mytest_security-policy.xml will
automatically be placed into the right directory (/etc/xen/acm-security/policies/).
If you run the tool on another system, then you need to copy the
resulting policy file into Domain-0 before continuing. See
Section 10.3.1 for naming conventions of security
policies.
Note: The support for __UNLABELED__ domains and
resources is meant to help transitioning from an uncontrolled
environment to a workload-protected environment by starting with
unlabeled domains and resources and then step-by-step labeling domains
and resources. Once all workloads are labeled, the __UNLABELED__
type can simply be removed from the Domain-0 label or from the policy
through a policy update. Section 10.3.5 will
show how unlabeled domains can be disabled by updating the
mytest policy at run-time.
The following command translates the source policy representation
into a format that can be loaded into Xen with sHype/ACM support,
activates the policy, and configures this policy for future boot
cycles into the boot sequence. Please refer to the xm
man page for further details:
(5) # xm setpolicy ACM mytest
Successfully set the new policy.
Supported security subsystems : ACM
Policy name : mytest
Policy type : ACM
Version of XML policy : 1.0
Policy configuration : loaded, activated for boot
Alternatively, if installing the policy fails (e.g., because it cannot identify the Xen boot entry), you can manually install the policy in 3 steps a-c.
(Alternatively to 5 - step a) Manually copy the policy binary file into the boot directory:
# cp /etc/xen/acm-security/policies/mytest.bin /boot/mytest.bin
(Alternatively to 5 - step b) Manually add a module line to your Xen boot entry so that grub loads this policy file during startup:
title XEN Devel with 2.6.18.8
kernel /xen.gz
module /vmlinuz-2.6.18.8-xen root=/dev/sda3 ro console=tty0
module /initrd-2.6.18.8-xen.img
module /mytest.bin
(Alternatively to 5 - step c) Reboot. Xen will choose the bootstrap label defined in the policy as Domain-0 label during reboot. After reboot, you can re-label Domain-0 at run-time, cf Section 10.2.7.
Assuming that command (5) succeeded or you followed the alternative instructions above, you should see the new policy and label appear when listing domains:
# xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 941 1 r----- 81.5 ACM:mytest:SystemManagement
If the security label at the end of the line says ``INACTIVE'' then the
security is not enabled. Verify the previous steps. Note: Domain-0 is
assigned a default label (see bootstrap policy attribute
explained in Section 10.3). All other domains must
be explicitly labeled, which we describe in detail below.
Unmanaged domains are started in Xen by using a configuration file. Please refer to Section 10.2.8 if you are using managed domains.
The following configuration file defines domain1
(Note: www.jailtime.org or www.xen.org might be good
places to look for example domU images):
# cat domain1.xm
kernel= "/boot/vmlinuz-2.6.18.8-xen"
memory = 128
name = "domain1"
vif = ['']
dhcp = "dhcp"
disk = ['file:/home/xen/dom_fc5/fedora.fc5.img,sda1,w', \
'file:/home/xen/dom_fc5/fedora.fc5.swap,sda2,w']
root = "/dev/sda1 ro xencons=tty"
Every domain must be associated with a security label before it can start
on sHype/Xen. Otherwise, sHype/Xen would not be able to enforce the policy
consistently. Our mytest policy is configured so that Xen
assigns a default label __UNLABELED__ to domains and resources that
have no label and supports them in a controlled manner. Since neither the domain,
nor the resources are (yet) labeled, this domain can start under the mytest
policy:
# xm create domain1.xm Using config file "./domain1.xm". Started domain domain1 # xm list --label Name ID Mem VCPUs State Time(s) Label domain1 1 128 1 -b---- 0.7 ACM:mytest:__UNLABELED__ Domain-0 0 875 1 r----- 84.6 ACM:mytest:SystemManagement
Please shutdown domain1 so that we can move it into the protection
domain of workload A-Bank.
# xm shutdown domain1 (wait some seconds until the domain has shut down) #xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 875 1 r----- 86.4 ACM:mytest:SystemManagement
We assume that the processing in domain1 contributes to the A-Bank workload.
We explore now how to transition this domain into the ``A-Bank'' workload-protection.
The following command prints all domain labels available in the active policy:
# xm labels A-Bank A-Bank.MarketAnalysis A-Bank.SecurityUnderwriting AutoCorp B-Bank SystemManagement __UNLABELED__
Now label domain1 with the A-Bank label and another domain2
with the B-Bank label. Please refer to the xm man page for
further information.
(6) # xm addlabel A-Bank dom domain1.xm
# xm addlabel B-Bank dom domain2.xm
Let us try to start the domain again:
# xm create domain1.xm Using config file "./domain1.xm". Error: VM's access to block device 'file:/home/xen/dom_fc5/fedora.fc5.img' denied
This error indicates that domain1, if started, would not be able to
access its image and swap files because they are not labeled. This
makes sense because to confine workloads, access of domains to
resources must be controlled. Otherwise, domains that are not allowed
to communicate or run simultaneously could share data through storage
resources.
xm labels type=res command to list available
resource labels. Let us assign the A-Bank resource label to the
domain1 image file representing /dev/sda1 and to its swap file:
(7) # xm addlabel A-Bank res \
file:/home/xen/dom_fc5/fedora.fc5.img
# xm addlabel A-Bank res \
file:/home/xen/dom_fc5/fedora.fc5.swap
The following command lists all labeled resources on the system, e.g., to lookup or verify the labeling:
# xm resources
file:/home/xen/dom_fc5/fedora.fc5.swap
type: ACM
policy: mytest
label: A-Bank
file:/home/xen/dom_fc5/fedora.fc5.img
type: ACM
policy: mytest
label: A-Bank
Starting domain1 will now succeed:
# xm create domain1.xm Using config file "./domain1.xm". Started domain domain1 # xm list --label Name ID Mem VCPUs State Time(s) Label domain1 3 128 1 -b---- 0.8 ACM:mytest:A-Bank Domain-0 0 875 1 r----- 90.9 ACM:mytest:SystemManagement
Currently, if a labeled resource is moved to another location, the
label must first be manually removed, and after the move re-attached
using the xm commands rmlabel and addlabel
respectively. Please see Section 10.4 for
further details.
(8) Label the resources of domain2 as B-Bank
but please do not start this domain yet.
We are about to demonstrate the sHype/Xen workload protection by verifying
domain1
with the A-Bank label is still running. While domain1 is running,
the run-time exclusion set of our policy implies that domain2 cannot
start because the label of domain1 includes the CHWALL type A-Bank
and the label of domain2 includes the CHWALL type B-Bank. The
run-time exclusion rule of our policy enforces that A-Bank and
B-Bank cannot run at the same time on the same hypervisor platform.
Once domain1 is stopped, saved, or migrated to another platform,
domain2 can start. Once domain2 is started, however,
domain1 can no longer start or resume on this system. When creating the
Chinese Wall types for the workload labels, the ezPolicy tool policy
translation component ensures that department workloads inherit all the
organization types (and with it any organization exclusions).
# xm list --label Name ID Mem VCPUs State Time(s) Label domain1 3 128 1 -b---- 0.8 ACM:mytest:A-Bank Domain-0 0 875 1 r----- 90.9 ACM:mytest:SystemManagement # xm create domain2.xm Using config file "./domain2.xm". Error: 'Domain in conflict set with running domains' # xm shutdown domain1 (wait some seconds until domain 1 is shut down) # xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 873 1 r----- 95.3 ACM:mytest:SystemManagement # xm create domain2.xm Using config file "./domain2.xm". Started domain domain2 # xm list --label Name ID Mem VCPUs State Time(s) Label domain2 5 164 1 -b---- 0.3 ACM:mytest:B-Bank Domain-0 0 839 1 r----- 96.4 ACM:mytest:SystemManagement # xm create domain1.xm Using config file "domain1.xm". Error: 'Domain in conflict with running domains' # xm shutdown domain2 # xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 839 1 r----- 97.8 ACM:mytest:SystemManagement
You can verify that domains with AutoCorp label can run together with domains labeled A-Bank or B-Bank.
domain1 with the B-Bank resource label. In a
real environment, the swap file must be sanitized (scrubbed/zeroed) before
it is reassigned to prevent data leaks from the A-Bank to the B-Bank workload
through the swap file.
We expect that domain1 will no longer start because it cannot access
this resource. This test checks the sharing abilities of domains, which are
defined by the Simple Type Enforcement Policy component.
# xm rmlabel res file:/home/xen/dom_fc5/fedora.fc5.swap
# xm addlabel B-Bank res file:/home/xen/dom_fc5/fedora.fc5.swap
# xm resources
file:/home/xen/dom_fc5/fedora.fc5.swap
type: ACM
policy: mytest
label: B-Bank
file:/home/xen/dom_fc5/fedora.fc5.img
type: ACM
policy: mytest
label: A-Bank
# xm create domain1.xm
Using config file "./domain1.xm".
Error:
VM's access to block device 'file:/home/xen/dom_fc5/fedora.fc5.swap' denied
The resource authorization checks are performed before the domain is actually started so that failures during the startup are prevented. A domain is only started if all the resources specified in its configuration are accessible.
In case no Domain-0 label is explicitly stated, we automatically assigned Domain-0
the SystemManagement label, which includes all STE (workload) types that
are known to the policy. In effect, the Domain-0 label authorizes the Xen system
to run only those workload types, whose STE types are included in the Domain-0
label. Hence, choosing the SystemManagement label for Domain-0 permits any
labeled domain to run. Resetting the label for Domain-0 at boot or run-time to
a label with a subset of the known STE workload types restricts which user domains
can run on this system. If Domain-0 is relabeled at run-time, then the new label
must at least include all STE types of those domains that are currently running.
The operation fails otherwise. This requirement ensures that the system remains
in a valid security configuration after re-labelling.
Restricting the Domain-0 authorization through the label creates a flexible policy-driven way to strongly partition the physical infrastructure and the workloads running on it. This partitioning will be automatically enforced during migration, start, or resume of domains and simplifies the security management considerably. Strongly competing workloads can be forced to run on separate physical infrastructure and become less depend on the domain isolation capabilities of the hypervisor.
First, we relabel the swap image back to A-Bank and then start up domain1:
# xm rmlabel res file:/home/xen/dom_fc5/fedora.fc5.swap # xm addlabel A-Bank res file:/home/xen/dom_fc5/fedora.fc5.swap # xm create domain1.xm Using config file "./domain1.xm". Started domain domain1 # xm list --label Name ID Mem VCPUs State Time(s) Label domain1 7 128 1 -b---- 0.7 ACM:mytest:A-Bank Domain-0 0 839 1 r----- 103.1 ACM:mytest:SystemManagement
The following command will restrict the Xen system to only run STE types included in the A-Bank label.
# xm addlabel A-Bank mgt Domain-0 Successfully set the label of domain 'Domain-0' to 'A-Bank'. # xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 839 1 r----- 103.7 ACM:mytest:A-Bank domain1 7 128 1 -b---- 0.7 ACM:mytest:A-Bank
In our example policy in Figure 10.4, this means that
only A-Bank domains and workloads (types) can run after the
successful completion of this command because the A-Bank label
includes only a single STE type, namely A-Bank. This command
fails if any running domain has an STE type in its label that is not
included in the A-Bank label.
If we now label a domain3 with AutoCorp, it cannot start because Domain-0 is
no longer authorized to run the workload type AutoCorp.
# xm addlabel AutoCorp dom domain3.xm (remember to label its resources, too) # xm create domain3.xm Using config file "./domain3.xm". Error: VM is not authorized to run. # xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 839 1 r----- 104.7 ACM:mytest:A-Bank domain1 7 128 1 -b---- 0.7 ACM:mytest:A-Bank
At this point, unlabeled domains cannot start either. Let domain4.xm describe an unlabeled domain, then trying to start domain4 will fail:
# xm getlabel dom domain4.xm Error: 'Domain not labeled' # xm create domain4.xm Using config file "./domain4.xm". Error: VM is not authorized to run.
Relabeling Domain-0 with the SystemManagement label will enable domain3 to start.
# xm addlabel SystemManagement mgt Domain-0 Successfully set the label of domain 'Domain-0' to 'SystemManagement'. # xm list --label Name ID Mem VCPUs State Time(s) Label domain1 7 128 1 -b---- 0.8 ACM:mytest:A-Bank Domain-0 0 839 1 r----- 106.6 ACM:mytest:SystemManagement # xm create domain3.xm Using config file "./domain3.xm". Started domain domain3 # xm list --label Name ID Mem VCPUs State Time(s) Label domain1 7 128 1 -b---- 0.8 ACM:mytest:A-Bank domain3 8 164 1 -b---- 0.3 ACM:mytest:AutoCorp Domain-0 0 711 1 r----- 107.6 ACM:mytest:SystemManagement
Xend has been extended with functionality to manage domains along with their
configuration information. Such domains are configured and started via Xen-API
calls. Since managed domains do not have an associated xm configuration file,
the existing addlabel command, which adds the security label into a
domain's configuration file, will not work for such managed domains.
Therefore, we have extended the xm addlabel and xm rmlabel
subcommands to enable adding security labels to and removing security
labels from managed domain configurations. The following example shows how
the A-Bank label can be assigned to the xend-managed
domain configuration of domain1. Removing labels from managed user
domain configurations works similarly.
Below, we show a dormant configuration of the managed domain1
with ID "-1" and state "-----" before labeling:
# xm list --label Name ID Mem VCPUs State Time(s) Label domain1 -1 128 1 ------ 0.0 ACM:mytest:__UNLABELED__ Domain-0 0 711 1 r----- 128.4 ACM:mytest:SystemManagement
Now we label the managed domain:
# xm addlabel A-Bank mgt domain1 Successfully set the label of the dormant domain 'domain1' to 'A-Bank'.
After labeling, you can see that the security label is part of the domain configuration:
# xm list --label Name ID Mem VCPUs State Time(s) Label domain1 -1 128 1 ------ 0.0 ACM:mytest:A-Bank Domain-0 0 711 1 r----- 129.7 ACM:mytest:SystemManagement
This command extension does not support relabeling of individual running user domains for several reasons. For one, because of the difficulty to revoke resources in cases where a running domain's new label does not permit access to resources that were accessible under the old label. Another reason is that changing the label of a single domain of a workload is rarely a good choice and will affect the workload isolation properties of the overall workload.
However, the name and contents of the label associated with running domains can be indirectly changed through a global policy change, which will update the whole workload consistently (domains and resources), cf. Section 10.3.5.
This section describes the sHype/Xen access control policy in detail. It gives enough information to enable the reader to write custom access control policies and to use the available Xen policy tools. The policy language is expressive enough to specify most symmetric access relationships between domains and resources efficiently.
The Xen access control policy consists of two policy components. The first component, called Simple Type Enforcement (STE) policy, controls the sharing between running domains, i.e., communication or access to shared resources. The second component, called Chinese Wall (CHWALL) policy, controls which domains can run simultaneously on the same virtualized platform. The CHWALL and STE policy components complement each other. The XML policy file includes all information needed by Xen to enforce those policies.
Figures 10.3 and 10.4 show the fully functional but very simple example Xen security policy that is created by ezPolicy as shown in Figure 10.2. The policy can distinguish the 6 workload types shown in lines 11-17 in Fig. 10.3. The whole XML Security Policy consists of four parts:
The Policy Header spans lines 4-8. It includes a date field and
defines the policy name mytest as well
as the version of the XML. It can also include optional fields that are
not shown and are for future use (see schema definition).
The policy name serves two purposes: First, it provides a unique name
for the security policy. This name is also exported by the Xen
hypervisor to the Xen management tools in order to ensure that both
the Xen hypervisor and Domain-0 enforce the same policy.
We plan to extend the policy name with a
digital fingerprint of the policy contents to better protect this
correlation. Second, it implicitly points the xm tools to the
location where the XML policy file is stored on the Xen system.
Replacing the colons in the policy name by slashes yields the local
path to the policy file starting from the global policy directory
/etc/xen/acm-security/policies. The last part of the policy
name is the prefix for the XML policy file name, completed by
-security_policy.xml. Our example policy with the name
mytest can be found in the XML policy file named
mytest-security_policy.xml that is stored under the global
policy directory. Another, preinstalled example policy named
example.test can be found in the test-security_policy.xml
under /etc/xen/acm-security/policies/example.
The Simple Type Enforcement (STE) policy controls which domains can communicate or share resources. This way, Xen can enforce confinement of workload types by confining the domains running those workload types and their resources. The mandatory access control framework enforces its policy when domains access intended communication or cooperation means (shared memory, events, shared resources such as block devices). It builds on top of the core hypervisor isolation, which restricts the ways of inter-communication to those intended means. STE does not protect or intend to protect from covert channels in the hypervisor or hardware; this is an orthogonal problem that can be mitigated by using the Run-time Exclusion rules described above or by fixing the problem leading to those covert channels in the core hypervisor or hardware platform.
Xen controls sharing between domains on the resource and domain level because this is the abstraction the hypervisor and its management understand naturally. While this is coarse-grained, it is also very reliable and robust and it requires minimal changes to implement mandatory access controls in the hypervisor. It enables platform- and operating system-independent policies as part of a layered security approach.
Lines 11-17 (cf Figure 10.3) define the Simple Type
Enforcement policy component. Essentially, they define the workload
type names SystemManagement, A-Bank,
AutoCorp etc. that are available in the STE policy component. The
policy rules are implicit: Xen permits two domains to communicate with
each other if and only if their security labels have at least one STE type in
common. Similarly, Xen permits a user domain to access a
resource if and only if the labels of the domain and the resource
have at least one STE workload type in common.
The Chinese Wall security policy interpretation of sHype enables users to prevent certain workloads from running simultaneously on the same hypervisor platform. Run-time Exclusion rules (RER), also called Conflict Sets or Anti-Collocation rules, define a set of workload types that are not permitted to run simultaneously on the same virtualized platform. Of all the workloads specified in a Run-time Exclusion rule, at most one type can run on the same hypervisor platform at a time. Run-time Exclusion Rules implement a less rigorous variant of the original Chinese Wall security component. They do not implement the *-property of the policy, which would require to restrict also types that are not part of an exclusion rule once they are running together with a type in an exclusion rule (http://www.gammassl.co.uk/topics/chinesewall.html provides more information on the original Chinese Wall policy).
Xen considers the ChineseWallTypes part of the label for the
enforcement of the Run-time Exclusion rules. It is illegal to define
labels including conflicting Chinese Wall types.
Lines 20-41 (cf Figure 10.3) define the Chinese Wall policy component. Lines 22-28 define the known Chinese Wall types, which coincide here with the STE types defined above. This usually holds if the criteria for sharing among domains and sharing of the hardware platform are the same. Lines 30-41 define one Run-time Exclusion rules, the first of which is depicted below:
31 <Conflict name="RER"> 32 <Type>A-Bank</Type> 33 <Type>B-Bank</Type> 34 <Type>__UNLABELED__</Type> 35 </Conflict>
Based on this rule, Xen enforces that only one of the types
A-Bank, B-Bank, or __UNLABELED__ will run
on a single hypervisor platform at a time. For example, once a domain assigned a
A-Bank workload type is started, domains with the
B-Bank type or unlabeled domains will be denied to start.
When the former domain stops and no other domains with the A-Bank
type are running, then domains with the B-Bank type or unlabeled domains
can start.
Xen maintains reference counts on each running workload type to keep track of which workload types are running. Every time a domain starts or resumes, the reference count on those Chinese Wall types that are referenced in the domain's label are incremented. Every time a domain is destroyed or saved, the reference counts of its Chinese Wall types are decremented. sHype in Xen fully supports migration and live-migration, which is subject to access control the same way as saving a domain on the source platform and resuming it on the destination platform.
Here are some reasons why users might want to restrict workloads or domains from sharing the system hardware simultaneously:
To enable Xen to associate domains with workload types running in them, each domain is assigned a security label that includes the workload types of the domain.
|
SecurityLabelTemplate (cf Figure 10.4) defines
the security labels that can be associated with domains and resources when
this policy is active (use the xm labels type=any command described in
Section 10.2.4 to list all available labels).
The domain labels include
Chinese Wall types while resource labels do not include Chinese Wall types.
The SubjectLabels policy section defines the labels that can be
assigned to domains. The VM label
A-Bank.SecurityUnderwriting in Figure 10.4)
associates the domain that carries it with the workload STE type
A-Bank.SecurityUnderwriting and with the CHWALL types A-Bank
and A-Bank.SecurityUnderwriting. The ezPolicy tool
assumes that any department workload will inherit any conflict set that
is specified for its organization, i.e., if B-Bank is running, not
only A-Bank but also all its departmental workloads are prevented
from running by this first run-time exclusion set. The separation of STE
and CHWALL types in the label definition ensures that
all departmental workloads are isolated from each other and from their generic
organization workloads, while they are sharing CHWALL types to
simplify the formulation of run-time exclusion sets.
The bootstrap attribute of the <SubjectLabels> XML node
in our example policy shown in Figure 10.4 names
the label SystemManagement as the label that Xen will assign
to Domain-0 at boot time (if this policy is installed as boot policy). The
label of Domain-0 can be persistently changed at run-time with the
addlabel command, which adds an overriding option to the grub.conf
boot entry (cf Section 10.2.7).
All user domains are assigned labels according to their domain configuration
(see Section 10.2.4 for examples of
how to label domains).
The ObjectLabels depicted in Figure 10.4 can be
assigned to resources when this policy is active.
In general, user domains should be assigned labels that have only a
single SimpleTypeEnforcement workload type. This way, workloads remain
confined even if user domains become rogue. Any domain that is
assigned a label with multiple STE types must be trusted to keep
information belonging to the different STE types separate (confined).
For example, Domain-0 is assigned the bootstrap label
SystemManagement, which includes all existing STE types.
Therefore, Domain-0 must take care not to enable unauthorized
information flow (eg. through block devices or virtual networking)
between domains or resources that are assigned different STE types.
Security administrators simply use the name of a label (specified in
the <Name> field) to associate a label with a domain (cf.
Section 10.2.4). The types inside the
label are used by the Xen access control enforcement. While the name
can be arbitrarily chosen (as long as it is unique), it is advisable
to choose the label name in accordance to the security types included.
Similarly, the STE and CHWALL types should be named according to the
workloads they represent. While the XML representation of the label
in the above example seems unnecessary flexible, labels in general
must be able to include multiple types.
We assume in the following example, that A-Bank.SecurityUnderwriting and
A-Bank.MarketAnalysis workloads use virtual disks that are provided
by a virtual I/O domain hosting a physical storage device and carrying
the following label:
<VirtualMachineLabel>
<Name>VIOServer</Name>
<SimpleTypeEnforcementTypes>
<Type>A-Bank</Type>
<Type>A-Bank.SecurityUnderwriting</Type>
<Type>A-Bank.MarketAnalysis</Type>
<Type>VIOServer</Type>
</SimpleTypeEnforcementTypes>
<ChineseWallTypes>
<Type>VIOServer</Type>
</ChineseWallTypes>
</VirtualMachineLabel>
This Virtual I/O domain (VIO) exports its virtualized disks by
communicating to all domains labeled with the
A-Bank.SecurityUnderwriting, the A-Bank, or the
A-Bank.MarketAnalysis label. This requires the
VIO domain to carry those STE types. In addition, this label includes a
new VIOServer type that can be used to restrict direct access to the
physical storage resource to the VIODomain.
In this example, the confinement of these A-Bank workloads depends on the
VIO domain that must keep the data of those different workloads separate.
The virtual disks are labeled as well to keep track of their assignments
to workload types (see Section 10.2.5
for labeling resources) and enforcement functions inside the VIO
domain must ensure that the labels of the domain mounting a virtual
disk and the virtual disk label share a common STE type. The VIO label
carrying its own VIOServer CHWALL type introduces the flexibility to
permit the trusted VIO server to run together with A-Bank.SecurityUnderwriting
or A-Bank.MarketAnalysis workloads.
Alternatively, a system that has two hard-drives does not need a VIO domain but can directly assign one hardware storage device to each of the workloads if the platform offers an IO-MMU, cf Section 9.2. Sharing hardware through virtualized devices is a trade-off between the amount of trusted code (size of the trusted computing base) and the amount of acceptable over-provisioning. This holds both for peripherals and for system platforms.
resetpolicy command:
# xm getpolicy
Supported security subsystems : ACM
Policy name : mytest
Policy type : ACM
Version of XML policy : 1.0
Policy configuration : loaded, activated for boot
# xm resetpolicy
Successfully reset the system's policy.
# xm getpolicy
Supported security subsystems : ACM
Policy name : DEFAULT
Policy type : ACM
Version of XML policy : 1.0
Policy configuration : loaded
# xm resources
file:/home/xen/dom_fc5/fedora.fc5.swap
type: INV_ACM
policy: mytest
label: A-Bank
file:/home/xen/dom_fc5/fedora.fc5.img
type: INV_ACM
policy: mytest
label: A-Bank
As the xm resources output shows, all resource labels have
invalidated type information but their semantics remain associated
with the resources so that they can later on either be relabeled
with semantically equivalent labels or sanitized and reused
(storage resources).
At this point, the system is in the same initial state as after configuring XSM and sHype/ACM and rebooting the system without a specific policy. No user domains can run.
# xm resetpolicy Successfully reset the system's policy. # xm setpolicy ACM example.test Successfully set the new policy. Supported security subsystems : ACM Policy name : example.test Policy type : ACM Version of XML policy : 1.0 Policy configuration : loaded, activated for boot # xm labels CocaCola PepsiCo SystemManagement VIO # xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 873 1 r----- 56.3 ACM:example.test:SystemManagement # xm resetpolicy Successfully reset the system's policy. # xm getpolicy Supported security subsystems : ACM Policy name : DEFAULT Policy type : ACM Version of XML policy : 1.0 Policy configuration : loaded # xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 873 1 r----- 57.2 ACM:DEFAULT:SystemManagement # xm setpolicy ACM mytest Successfully set the new policy. Supported security subsystems : ACM Policy name : mytest Policy type : ACM Version of XML policy : 1.0 Policy configuration : loaded, activated for boot # xm labels A-Bank A-Bank.MarketAnalysis A-Bank.SecurityUnderwriting AutoCorp B-Bank SystemManagement __UNLABELED__ # xm list --label Name ID Mem VCPUs State Time(s) Label Domain-0 0 873 1 r----- 58.0 ACM:mytest:SystemManagement
The described way of changing policies by resetting the existing policy is useful for testing different policies. For real deployment environments, a policy update as described in the following section is more appropriate and can be applied seamlessly at run-time while user domains are running.
Once an ACM security policy is activated (loaded into the Xen hypervisor), the policy may be updated at run-time without the need to re-boot the system. The XML update-policy contains several additional information fields that are required to safely link the new policy contents to the old policy and ensure a consistent transformation of the system security state from the old to the new policy. Those additional fields are required for policies that are updating an existing policy at run-time.
The major benefit of policy updates is the ability to add, delete, or rename workload types, labels, and conflict sets (run-time exclusion rules) to accommodate changes in the managed virtual environment without the need to reboot the Xen system. When a new policy renames labels of the current policy, the labels attached to resources and domains are automatically updated during a successful policy update.
We have manually crafted an update policy for the mytest
security policy and stored it in the file mytest_update-security_policy.xml
in the policies directory. We will discuss this policy in detail before
using it to update a running sHype/Xen system. The following figures contain
the whole contents of the update policy file.
Figure 10.5 shows the policy
header of an update-policy and the new FromPolicy XML
node. For the policy update to succeed, the policy name and the
policy version fields of the FromPolicy XML node must
exactly match those of the currently enforced policy. This
ensures a controlled update path of the policy.
The version number of the new policy, which is shown in the
node following the Date node, must be a logical increment
to the current policy's version. Therefore at least the minor
number of the policy version must be incremented. This ensures
that a policy update is applied only to exactly the policy for
which this update was created and minimizes unforseen side-effects
of policy updates.
__UNLABELED__ type is removed
to disable support for running unlabeled domains. Additionally,
we have renamed the two A-Bank department types with
abbreviated names A-Bank.SU and A-Bank.MA. You
can also see how those type names are
consistently changed within the conflict set definition.
In the same way, new types can be introduced and new conflict sets can be defined by simply adding the types or conflict sets to the update policy.
A policy update also enables the renaming of virtual machine and
resource labels. Linking the old label name with the new label
name is achieved through the from attribute in the
VirtualMachineLabel or ResourceLabel nodes in the
update-policy. Figure 10.7 shown how subject
and resource labels
are updated from their old name A-Bank.SecurityUnterwriting
to their new name A-Bank.SU using the from attribute.
|
The updated label definition also includes a new label A-Bank-WL
that includes all STE types related to A-Bank. Its CHWALL type
is SystemManagement. This indicates that this label is designed
as Domain-0 label. A Xen system can be restricted to only run A-Bank
related workloads by relabeling Domain-0 with the A-Bank-WL label.
We assume that the update-policy shown in Figures 10.5, 10.6 and 10.7 is stored in the XML file mytest_update-security_policy.xml located in the ACM policy directory. See Section 10.3.1 for information about policy names and locations.
The following xm setpolicy command updates the active ACM
security policy at run-time.
# xm list --label
Name ID Mem VCPUs State Time(s) Label
domain1 2 128 1 -b---- 0.6 ACM:mytest:A-Bank
domain4 3 164 1 -b---- 0.3 ACM:mytest:A-Bank.SecurityUnderwriting
Domain-0 0 711 1 r----- 71.8 ACM:mytest:SystemManagement
# xm resources
file:/home/xen/dom_fc5/fedora.fc5.swap
type: ACM
policy: mytest
label: A-Bank
file:/home/xen/dom_fc5/fedora.fc5.img
type: ACM
policy: mytest
label: A-Bank
# xm setpolicy ACM mytest_update
Successfully set the new policy.
Supported security subsystems : ACM
Policy name : mytest
Policy type : ACM
Version of XML policy : 1.1
Policy configuration : loaded, activated for boot
# xm list --label
Name ID Mem VCPUs State Time(s) Label
domain1 2 128 1 -b---- 0.7 ACM:mytest:A-Bank
domain4 3 164 1 -b---- 0.3 ACM:mytest:A-Bank.SU
Domain-0 0 711 1 r----- 72.8 ACM:mytest:SystemManagement
# xm labels
A-Bank
A-Bank-WL
A-Bank.MA
A-Bank.SU
AutoCorp
B-Bank
# xm resources
file:/home/xen/dom_fc5/fedora.fc5.swap
type: ACM
policy: mytest
label: A-Bank
file:/home/xen/dom_fc5/fedora.fc5.img
type: ACM
policy: mytest
label: A-Bank
After successful completion of this command, xm list --label
shows that the labels of running domains changed to their new names.
xm labels shows that new labels A-Bank.SU and A-Bank.AM
are now available in the policy. The resource labels remain valid after
the successful update as xm resources confirms.
The setpolicy command fails if the new policy is inconsistent
with the current one or the policy is inconsistent internally (e.g., types
are renamed in the type definition but not in the label definition part of
the policy). In this case, the old policy remains active.
After relabeling Domain-0 with the new A-Bank-WL label, we can no
longer run domains labeled B-Bank or AutoCorp since their
STE types are not a subset of the new Domain-0 label.
# xm addlabel A-Bank-WL mgt Domain-0 Successfully set the label of domain 'Domain-0' to 'A-Bank-WL'. # xm list --label Name ID Mem VCPUs State Time(s) Label domain1 2 128 1 -b---- 0.8 ACM:mytest:A-Bank Domain-0 0 711 1 r----- 74.5 ACM:mytest:A-Bank-WL domain4 3 164 1 -b---- 0.3 ACM:mytest:A-Bank.SU # xm getlabel dom domain3.xm policytype=ACM,policy=mytest,label=AutoCorp # xm create domain3.xm Using config file "./domain3.xm". Error: VM is not authorized to run. # xm addlabel SystemManagement mgt Domain-0 Successfully set the label of domain 'Domain-0' to 'SystemManagement'. # xm list --label Name ID Mem VCPUs State Time(s) Label domain1 2 128 1 -b---- 0.8 ACM:mytest:A-Bank domain4 3 164 1 -b---- 0.3 ACM:mytest:A-Bank.SU Domain-0 0 709 1 r----- 76.4 ACM:mytest:SystemManagement # xm create domain3.xm Using config file "./domain3.xm". Started domain domain3 # xm list --label Name ID Mem VCPUs State Time(s) Label domain1 2 128 1 -b---- 0.8 ACM:mytest:A-Bank domain4 3 164 1 -b---- 0.3 ACM:mytest:A-Bank.SU domain3 4 164 1 -b---- 0.3 ACM:mytest:AutoCorp Domain-0 0 547 1 r----- 77.5 ACM:mytest:SystemManagement
In the same manner, you can add new labels to support new workloads and add, delete, or rename workload types (STE and/or CHWALL types) simply by changing the composition of labels. Another use case is to add new workload types to the current Domain-0 label to enable them to run. Conflict sets (run-time exclusion rules) can be simply omitted or added. The policy and label changes become active at once and new workloads can be run in protected mode without rebooting the Xen system.
In all these cases, if any running user domain would-under the new policy-not be allowed to run or would not be allowed to access any of the resources it currently uses, then the policy update is rejected. In this case, you can stop domains that conflict with the new policy and update the policy afterwards. The old policy remains active until a policy update succeeds or Xen is re-booted into a new policy.
ezPolicy GUI tool - start writing policies
xensec_gen tool - refine policies created with ezPolicy
We use the ezPolicy tool in
Section 10.2.2 to quickly create a workload
protection policy. If desired, the resulting XML policy file can be
loaded into the xensec_gen tool to refine it. It can also be
directly edited using an XML editor. Any XML policy file is verified
against the security policy schema when it is translated (see
Subsection 10.2.3).
The sHype/ACM configuration for Xen is work in progress. There is ongoing work for protecting virtualized resources and planned and ongoing work for protecting access to remote resources and domains. The following sections describe limitations of some of the areas into which access control is being extended.
Enforcing the security policy across multiple hypervisor systems and on access to remote shared resources is work in progress. Extending access control to new types of resources is ongoing work (e.g. network storage).
On a single Xen system, information about the association of resources
and security labels is stored in
/var/lib/xend/security/policies/resource_labels. This file relates
a full resource path with a security label. This association is weak
and will break if resources are moved or renamed without adapting the
label file. Improving the protection of label-resource relationships
is ongoing work.
Controlling resource usage and enforcing resource limits in general is ongoing work in the Xen community.
Labels on domains are enforced during domain migration and the destination hypervisor will ensure that the domain label is valid and the domain is permitted to run (considering the Chinese Wall policy rules) before it accepts the migration. However, the network between the source and destination hypervisor as well as both hypervisors must be trusted. Architectures and prototypes exist that both protect the network connection and ensure that the hypervisors enforce access control consistently but patches are not yet available for the main stream.
The sHype access control aims at system independent security policies. It builds on top of the core hypervisor isolation. Any covert channels that exist in the core hypervisor or in the hardware (e.g., shared processor cache) will be inherited. If those covert channels are not the result of trade-offs between security and other system properties, then they are most effectively minimized or eliminated where they are caused. sHype offers however some means to mitigate their impact, e.g., run-time exclusion rules (cf Section 10.2.2) or limiting the system authorization (cf Section 10.2.7).
This chapter describes the build- and boot-time options which may be used to tailor your Xen system.
Top-level configuration is achieved by editing one of two files: Config.mk and Makefile.
The former allows the overall build target architecture to be specified. You will typically not need to modify this unless you are cross-compiling or if you wish to build a non-PAE Xen system. Additional configuration options are documented in the Config.mk file.
The top-level Makefile is chiefly used to customize the set of kernels built. Look for the line:
KERNELS ?= linux-2.6-xen0 linux-2.6-xenU
Allowable options here are any kernels which have a corresponding build configuration file in the buildconfigs/ directory.
Xen provides a number of build-time options which should be set as environment variables or passed on make's command-line.
These options are used to configure Xen's behaviour at runtime. They should be appended to Xen's command line, either manually or by editing grub.conf.
In addition, the following options may be specified on the Xen command line. Since domain 0 shares responsibility for booting the platform, Xen will automatically propagate these options to its command line. These options are taken from Linux's command-line syntax with unchanged semantics.
In addition to the standard Linux kernel boot options, we support:
| `xencons=off': disable virtual console |
| `xencons=tty': attach console to /dev/tty1 (tty0 at boot-time) |
| `xencons=ttyS': attach console to /dev/ttyS0 |
| `xencons=xvc': attach console to /dev/xvc0 |
If you have questions that are not answered by this manual, the sources of information listed below may be of interest to you. Note that bug reports, suggestions and contributions related to the software (or the documentation) should be sent to the Xen developers' mailing list (address below).
For developers interested in porting operating systems to Xen, the Xen Interface Manual is distributed in the docs/ directory of the Xen source distribution.
The official Xen web site can be found at:
http://www.xensource.com
This contains links to the latest versions of all online documentation, including the latest version of the FAQ.
Information regarding Xen is also available at the Xen Wiki at
http://wiki.xensource.com/xenwiki/The Xen project uses Bugzilla as its bug tracking system. You'll find the Xen Bugzilla at http://bugzilla.xensource.com/bugzilla/.
There are several mailing lists that are used to discuss Xen related topics. The most widely relevant are listed below. An official page of mailing lists and subscription information can be found at
http://lists.xensource.com/
Xen supports guest domains running unmodified guest operating systems using virtualization extensions available on recent processors. Currently processors featuring the Intel Virtualization Extension (Intel-VT) or the AMD extension (AMD-V) are supported. The technology covering both implementations is called HVM (for Hardware Virtual Machine) in Xen. More information about the virtualization extensions are available on the respective websites: http://www.intel.com/technology/computing/vptech
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/24593.pdf
The following packages need to be installed in order to build Xen with HVM support. Some Linux distributions do not provide these packages by default.
| Package | Description |
| dev86 | The dev86 package provides an assembler and linker for real mode 80x86 instructions. You need to have this package installed in order to build the BIOS code which runs in (virtual) real mode.
If the dev86 package is not available on the x86_64 distribution, you can install the i386 version of it. The dev86 rpm package for various distributions can be found at http://www.rpmfind.net/linux/rpm2html/search.php?query=dev86&submit=Search |
| SDL-devel, SDL | Simple DirectMedia Layer (SDL) is another way of virtualizing the unmodified guest console. It provides an X window for the guest console.
If the SDL and SDL-devel packages are not installed by default on the build system, they can be obtained from http://www.rpmfind.net/linux/rpm2html/search.php?query=SDL&submit=Search http://www.rpmfind.net/linux/rpm2html/search.php?query=SDL-devel&submit=Search |
The Xen installation includes a sample configuration file, /etc/xen/xmexample.hvm. There are comments describing all the options. In addition to the common options that are the same as those for paravirtualized guest configurations, HVM guest configurations have the following settings:
| Parameter | Description |
| kernel | The HVM firmware loader, /usr/lib/xen/boot/hvmloader |
| builder | The domain build function. The HVM domain uses the 'hvm' builder. |
| acpi | Enable HVM guest ACPI, default=1 (enabled) |
| apic | Enable HVM guest APIC, default=1 (enabled) |
| pae | Enable HVM guest PAE, default=1 (enabled) |
| hap | Enable hardware-assisted paging support, such as AMD-V's nested paging or Intel®VT's extended paging. If available, Xen will use hardware-assisted paging instead of shadow paging for this guest's memory management. |
| vif | Optionally defines MAC address and/or bridge for the network interfaces. Random MACs are assigned if not given. type=ioemu means ioemu is used to virtualize the HVM NIC. If no type is specified, vbd is used, as with paravirtualized guests. |
| disk | Defines the disk devices you want the domain to have access to, and what you want them accessible as. If using a physical device as the HVM guest's disk, each disk entry is of the form
phy:UNAME,ioemu:DEV,MODE, where UNAME is the host device file, DEV is the device name the domain will see, and MODE is r for read-only, w for read-write. ioemu means the disk will use ioemu to virtualize the HVM disk. If not adding ioemu, it uses vbd like paravirtualized guests. If using disk image file, its form should be like file:FILEPATH,ioemu:DEV,MODE Optical devices can be emulated by appending cdrom to the device type ',hdc:cdrom,r' If using more than one disk, there should be a comma between each disk entry. For example: disk = ['file:/var/images/image1.img,ioemu:hda,w', 'phy:hda1,hdb1,w', 'file:/var/images/install1.iso,hdc:cdrom,r'] |
| boot | Boot from floppy (a), hard disk (c) or CD-ROM (d). For example, to boot from CD-ROM and fallback to HD, the entry should be:
boot='dc' |
| device_model | The device emulation tool for HVM guests. This parameter should not be changed. |
| sdl | Enable SDL library for graphics, default = 0 (disabled) |
| vnc | Enable VNC library for graphics, default = 1 (enabled) |
| vncconsole | Enable spawning of the vncviewer (only valid when vnc=1), default = 0 (disabled)
If vnc=1 and vncconsole=0, user can use vncviewer to manually connect HVM from remote. For example: vncviewer domain0_IP_address:HVM_domain_id |
| serial | Enable redirection of HVM serial output to pty device |
| usb | Enable USB support without defining a specific USB device. This option defaults to 0 (disabled) unless the option usbdevice is specified in which case this option then defaults to 1 (enabled). |
| usbdevice | Enable USB support and also enable support for the given
device. Devices that can be specified are mouse (a PS/2 style
mouse), tablet (an absolute pointing device) and
host:id1:id2 (a physical USB device on the host machine whose
ids are id1 and id2). The advantage
of tablet is that Windows guests will automatically recognize
and support this device so specifying the config line
will create a mouse that works transparently with Windows guests under VNC. Linux doesn't recognize the USB tablet yet so Linux guests under VNC will still need the Summagraphics emulation. Details about mouse emulation are provided in section A.4.3. |
| localtime | Set the real time clock to local time [default=0, that is, set to UTC]. |
| soundhw | Enable sound card support and specify the hardware to emulate. Values can be sb16, es1370 or all. Default is none. |
| full-screen | Start in full screen. |
| nographic | Another way to redirect serial output. If enabled, no 'sdl' or 'vnc' can work. Not recommended. |
# dd if=/dev/zero of=hd.img bs=1M count=0 seek=1024
Install Xen and create HVM with the original image file with booting from CD-ROM. Then it is just like a normal Linux OS installation. The HVM configuration file should have a stanza for the CD-ROM as well as a boot device specification:
disk=['file:/var/images/your-hd.img,hda,w', ',hdc:cdrom,r' ] boot='d'
If this method does not succeed, you can choose the following method of copying an installed Linux OS into an image file.
# losetup /dev/loop0 hd.img
# fdisk -b 512 -C 4096 -H 16 -S 32 /dev/loop0
press 'n' to add new partition
press 'p' to choose primary partition
press '1' to set partition number
press "Enter" keys to choose default value of "First Cylinder" parameter.
press "Enter" keys to choose default value of "Last Cylinder" parameter.
press 'w' to write partition table and exit
# losetup -d /dev/loop0
The losetup option -o 16384 skips the partition table in the image file. It is the number of sectors times 512. We need /dev/loop because grub is expecting a disk device name, where name represents the entire disk and name1 represents the first partition.
# lomount -t ext3 -diskimage hd.img -partition 1 /mnt/guest
# cp -ax /var/guestos/{root,dev,var,etc,usr,bin,sbin,lib} /mnt/guest
# mkdir /mnt/guest/{proc,sys,home,tmp}
# vim /mnt/guest/etc/fstab
/dev/hda1 / ext3 defaults 1 1
none /dev/pts devpts gid=5,mode=620 0 0
none /dev/shm tmpfs defaults 0 0
none /proc proc defaults 0 0
none /sys sysfs efaults 0 0
Now, the guest OS image hd.img is ready. You can also reference http://free.oszoo.org for quickstart images. But make sure to install the boot loader.
disk = [ 'file:/var/images/min-el3-i386.img,hda,w' ]
replacing min-el3-i386.img with the name of the guest OS image file you just made.
# xend start
# xm create /etc/xen/hvmguest.hvm
In the default configuration, VNC is on and SDL is off. Therefore VNC windows will open when HVM guests are created. If you want to use SDL to create HVM guests, set sdl=1 in your HVM configuration file. You can also turn off VNC by setting vnc=0.
To deal with these mouse issues there are 4 different mouse emulations available from the HVM device model:
Linux configuration.
First, configure the GPM service to use the Summagraphics tablet. This can vary between distributions but, typically, all that needs to be done is modify the file /etc/sysconfig/mouse to contain the lines:
MOUSETYPE="summa"
XMOUSETYPE="SUMMA"
DEVICE=/dev/ttyS1
and then restart the GPM daemon.
Next, modify the X11 config /etc/X11/xorg.conf to support the Summgraphics tablet by replacing the input device stanza with the following:
Section "InputDevice"
Identifier "Mouse0"
Driver "summa"
Option "Device" "/dev/ttyS1"
Option "InputFashion" "Tablet"
Option "Mode" "Absolute"
Option "Name" "EasyPen"
Option "Compatible" "True"
Option "Protocol" "Auto"
Option "SendCoreEvents" "on"
Option "Vendor" "GENIUS"
EndSection
Restart X and the X cursor should now properly track the VNC pointer.
Windows configuration.
Get the file http://www.cad-plan.de/files/download/tw2k.exe and execute that file on the guest, answering the questions as follows:
usbdevice='mouse'
usbdevice='tablet'
Linux configuration.
Unfortunately, there is no GPM support for the USB tablet at this point in time. If you intend to use a GPM pointing device under VNC you should configure the guest for Summagraphics emulation.
Support for X11 is available by following
the instructions at
http://stz-softwaretechnik.com/~ke/touchscreen/evtouch.html
with one minor change.
The
xorg.conf
given in those instructions
uses the wrong values for the X & Y minimums and maximums,
use the following config stanza instead:
Section "InputDevice"
Identifier "Tablet"
Driver "evtouch"
Option "Device" "/dev/input/event2"
Option "DeviceName" "touchscreen"
Option "MinX" "0"
Option "MinY" "0"
Option "MaxX" "32256"
Option "MaxY" "32256"
Option "ReportingMode" "Raw"
Option "Emulate3Buttons"
Option "Emulate3Timeout" "50"
Option "SendCoreEvents" "On"
EndSection
Windows configuration.
Just enabling the USB tablet in the guest's configuration file is sufficient, Windows will automatically recognize and configure device drivers for this pointing device.
usbdevice='mouse'
to the configuration file.
usbdevice='tablet'
to the configuration file.
usbdevice='host:vid:pid'
into the the configuration file.A.1vid and pid are a product id and vendor id that uniquely identify the USB device. These ids can be identified in two ways:
usb=1
then executing the command
info usbhost
in the control window
will display a list of all
usb devices and their ids.
For example,
this output:
Device 1.3, speed 1.5 Mb/s
Class 00: USB device 04b3:310b
was created from a USB mouse with
vendor id
04b3
and product id
310b.
This device could be made available
to the HVM guest by including the
config file entry
usbdevice='host:04be:310b'
It is also possible to enable access to a USB device dynamically through the control window. The control window command
usb_add host:vid:pid
will also allow access to a
USB device with vendor id
vid
and product id
pid.
T: Bus=01 Lev=01 Prnt=01 Port=01 Cnt=02 Dev#= 3 Spd=1.5 MxCh= 0 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs= 1 P: Vendor=04b3 ProdID=310b Rev= 1.60 C:* #Ifs= 1 Cfg#= 1 Atr=a0 MxPwr=100mA I: If#= 0 Alt= 0 #EPs= 1 Cls=03(HID ) Sub=01 Prot=02 Driver=(none) E: Ad=81(I) Atr=03(Int.) MxPS= 4 Ivl=10msNote that the P: line correctly identifies the vendor id and product id for this mouse as 04b3:310b.
Going back to the USB mouse as an example, if lsmod gives the output:
Module Size Used by usbmouse 4128 0 usbhid 28996 0 uhci_hcd 35409 0
then the USB mouse is being used by the Dom0 kernel and is not available to the guest. Executing the command rmmod usbhidA.2will remove the USB mouse driver from the Dom0 kernel and the mouse will now be accessible by the HVM guest.
Be aware the the Linux USB hotplug system will reload the drivers if a USB device is removed and plugged back in. This means that just unloading the driver module might not be sufficient if the USB device is removed and added back. A more reliable technique is to first rmmod the driver and then rename the driver file in the /lib/modules directory, just to make sure it doesn't get reloaded.
poweroff
in the HVM guest's console, for Windows use Start -> Shutdown first to prevent data loss. Depending on the configuration the guest will be automatically destroyed, otherwise execute the command
xm destroy vmx_guest_id
at the Domain0 console.
Ctrl+Alt+2 switches from guest VGA window to the control window. Typing help shows the control commands help. For example, 'q' is the command to destroy the HVM guest.
Ctrl+Alt+1 switches back to HVM guest's VGA.
Ctrl+Alt+3 switches to serial port output. It captures serial output from the HVM guest. It works only if the HVM guest was configured to use the serial port.
Xen optionally supports virtual networking for domains using vnets. These emulate private LANs that domains can use. Domains on the same vnet can be hosted on the same machine or on separate machines, and the vnets remain connected if domains are migrated. Ethernet traffic on a vnet is tunneled inside IP packets on the physical network. A vnet is a virtual network and addressing within it need have no relation to addressing on the underlying physical network. Separate vnets, or vnets and the physical network, can be connected using domains with more than one network interface and enabling IP forwarding or bridging in the usual way.
Vnet support is included in xm and xend:
# xm vnet-create <config>creates a vnet using the configuration in the file
<config>.
When a vnet is created its configuration is stored by xend and the vnet persists until it is
deleted using
# xm vnet-delete <vnetid>The vnets xend knows about are listed by
# xm vnet-listMore vnet management commands are available using the vn tool included in the vnet distribution.
The format of a vnet configuration file is
(vnet (id <vnetid>)
(bridge <bridge>)
(vnetif <vnet interface>)
(security <level>))
White space is not significant. The parameters are:
<vnetid>: vnet id, the 128-bit vnet identifier. This can be given
as 8 4-digit hex numbers separated by colons, or in short form as a single 4-digit hex number.
The short form is the same as the long form with the first 7 fields zero.
Vnet ids must be non-zero and id 1 is reserved.
<bridge>: the name of a bridge interface to create for the vnet. Domains
are connected to the vnet by connecting their virtual interfaces to the bridge.
Bridge names are limited to 14 characters by the kernel.
<vnetif>: the name of the virtual interface onto the vnet (optional). The
interface encapsulates and decapsulates vnet traffic for the network and is attached
to the vnet bridge. Interface names are limited to 14 characters by the kernel.
<level>: security level for the vnet (optional). The level may be one of
none: no security (default). Vnet traffic is in clear on the network.
auth: authentication. Vnet traffic is authenticated using IPSEC
ESP with hmac96.
conf: confidentiality. Vnet traffic is authenticated and encrypted
using IPSEC ESP with hmac96 and AES-128.
(vnet (id 97) (bridge vnet97) (vnetif vnif97)
(security none))
Then xm vnet-create vnet97.sxp will define a vnet with id 97 and no security.
The bridge for the vnet is called vnet97 and the virtual interface for it is vnif97.
To add an interface on a domain to this vnet set its bridge to vnet97
in its configuration. In Python:
vif="bridge=vnet97"In sxp:
(dev (vif (mac aa:00:00:01:02:03) (bridge vnet97)))Once the domain is started you should see its interface in the output of brctl show under the ports for vnet97.
To get best performance it is a good idea to reduce the MTU of a domain's interface onto a vnet to 1400. For example using ifconfig eth0 mtu 1400 or putting MTU=1400 in ifcfg-eth0. You may also have to change or remove cached config files for eth0 under /etc/sysconfig/networking. Vnets work anyway, but performance can be reduced by IP fragmentation caused by the vnet encapsulation exceeding the hardware MTU.
(network-script network-vnet)This script insmods the module and calls the network-bridge script.
The vnet code is not compiled and installed by default. To compile the code and install on the current system use make install in the root of the vnet source tree, tools/vnet. It is also possible to install to an installation directory using make dist. See the Makefile in the source for details.
The vnet module creates vnet interfaces vnif0002, vnif0003 and vnif0004 by default. You can test that vnets are working by configuring IP addresses on these interfaces and trying to ping them across the network. For example, using machines hostA and hostB:
hostA# ifconfig vnif0004 192.0.2.100 up hostB# ifconfig vnif0004 192.0.2.101 up hostB# ping 192.0.2.100
The vnet implementation uses IP multicast to discover vnet interfaces, so all machines hosting vnets must be reachable by multicast. Network switches are often configured not to forward multicast packets, so this often means that all machines using a vnet must be on the same LAN segment, unless you configure vnet forwarding.
You can test multicast coverage by pinging the vnet multicast address:
# ping -b 224.10.0.1You should see replies from all machines with the vnet module running. You can see if vnet packets are being sent or received by dumping traffic on the vnet UDP port:
# tcpdump udp port 1798
If multicast is not being forwarded between machines you can configure multicast forwarding using vn. Suppose we have machines hostA on 192.0.2.200 and hostB on 192.0.2.211 and that multicast is not forwarded between them. We use vn to configure each machine to forward to the other:
hostA# vn peer-add hostB hostB# vn peer-add hostAMulticast forwarding needs to be used carefully - you must avoid creating forwarding loops. Typically only one machine on a subnet needs to be configured to forward, as it will forward multicasts received from other machines on the subnet.