libvir.html

<html>
<head>
  <meta http-equiv="Content-Type" content="">
  <title>Libvirt the virtualization API</title>
</head>

<body bgcolor="#ffffff">
<h1 align="center">Libvirt the virtualization API</h1>

<h1>Note: this is the flat content of the <a href="index.html">web
site</a></h1>

<h1 style="text-align: center">libvirt</h1>

<h3>what is <span class="style1">libvirt?</span></h3>

<p>Libvirt is a C toolkit to interact with the virtualization capabilities
of recent versions of Linux (and other OSes). It is free software available
under the <a href="http://www.opensource.org/licenses/lgpl-license.html">GNU
Lesser General Public License</a>. Virtualization of the Linux Operating
System means the ability to run multiple instances of Operating Systems
concurently on a single hardware system where the basic resources are driven
by a Linux instance. The library aim at providing long term stable C API
initially for the <a
href="http://www.cl.cam.ac.uk/Research/SRG/netos/xen/index.html">Xen
paravirtualization</a> but should be able to integrate other
virtualization mechanisms, it currently also support <a
href="http://fabrice.bellard.free.fr/qemu/">QEmu</a> and <a
href="http://kvm.qumranet.com/">KVM</a>.</p>

<h2><a name="News">Releases</a></h2>

<p>Here is the list of official releases, however since it is early on in the
development of libvirt, it is preferable when possible to just use the <a
href="downloads.html">CVS version or snapshot</a>, contact the mailing list
and check the <a href="ChangeLog.html">ChangeLog</a> to gauge progresses.</p>

<h3>0.2.2: Apr 17 2007</h3>
<ul>
  <li>Documentation: fix errors due to Amaya (with Simon Hernandez), 
      virsh uses kB not bytes (Atsushi SAKAI), add command line help to
      qemud (Richard Jones), xenUnifiedRegister docs (Atsushi SAKAI),
      strings typos (Nikolay Sivov), ilocalization probalem raised by 
      Thomas Canniot</li>
  <li>Bug fixes: virsh memory values test (Masayuki Sunou), operations without
      libvirt_qemud (Atsushi SAKAI), fix spec file (Florian La Roche, Jeremy
      Katz, Michael Schwendt),
      direct hypervisor call (Atsushi SAKAI), buffer overflow on qemu
      networking command (Daniel Berrange), buffer overflow in quemud (Daniel
      Berrange), virsh vcpupin bug (Masayuki Sunou), host PAE detections
      and strcuctures size (Richard Jones), Xen PAE flag handling (Daniel
      Berrange), bridged config configuration (Daniel Berrange), erroneous
      XEN_V2_OP_SETMAXMEM value (Masayuki Sunou), memory free error (Mark
      McLoughlin), set VIR_CONNECT_RO on read-only connections (S.Sakamoto),
      avoid memory explosion bug (Daniel Berrange), integer overflow 
      for qemu CPU time (Daniel Berrange), QEMU binary path check (Daniel
      Berrange)</li>
  <li>Cleanups: remove some global variables (Jim Meyering), printf-style
      functions checks (Jim Meyering), better virsh error messages, increase
      compiler checkings and security (Daniel Berrange), virBufferGrow usage
      and docs, use calloc instead of malloc/memset, replace all sprintf by
      snprintf, avoid configure clobbering user's CTAGS (Jim Meyering), 
      signal handler error cleanup (Richard Jones), iptables internal code
      claenup (Mark McLoughlin), unified Xen driver (Richard Jones),
      cleanup XPath libxml2 calls, IPTables rules tightening (Daniel
      Berrange), </li>
  <li>Improvements: more regression tests on XML (Daniel Berrange), Python
      bindings now generate exception in error cases (Richard Jones),
      Python bindings for vir*GetAutoStart (Daniel Berrange),
      handling of CD-Rom device without device name (Nobuhiro Itou),
      fix hypervisor call to work with Xen 3.0.5 (Daniel Berrange),
      DomainGetOSType for inactive domains (Daniel Berrange), multiple boot
      devices for HVM (Daniel Berrange), 
      </li>
</ul>
<h3>0.2.1: Mar 16 2007</h3>
<ul>
  <li>Various internal cleanups (Richard Jones,Daniel Berrange,Mark McLoughlin)</li>
  <li>Bug fixes: libvirt_qemud daemon path (Daniel Berrange), libvirt
      config directory (Daniel Berrange and Mark McLoughlin), memory leak
      in qemud (Mark), various fixes on network support (Mark), avoid Xen
      domain zombies on device hotplug errors (Daniel Berrange), various
      fixes on qemud (Mark), args parsing (Richard Jones), virsh -t argument
      (Saori Fukuta), avoid virsh crash on TAB key (Daniel Berrange), detect
      xend operation failures (Kazuki Mizushima), don't listen on null socket
      (Rich Jones), read-only socket cleanup (Rich Jones), use of vnc port 5900
      (Nobuhiro Itou), assorted networking fixes (Daniel Berrange), shutoff and
      shutdown mismatches (Kazuki Mizushima), unlimited memory handling
      (Atsushi SAKAI), python binding fixes (Tatsuro Enokura)</li>
  <li>Build and portability fixes: IA64 fixes (Atsushi SAKAI), dependancies
      and build (Daniel Berrange), fix xend port detection (Daniel
      Berrange), icompile time warnings (Mark), avoid const related
      compiler warnings (Daniel Berrange), automated builds (Daniel
      Berrange), pointer/int mismatch (Richard Jones), configure time
      selection of drivers, libvirt spec hacking (Daniel Berrange)</li>
  <li>Add support for network autostart and init scripts (Mark McLoughlin)</li>
  <li>New API virConnectGetCapabilities() to detect the virtualization 
    capabilities of a host (Richard Jones)</li>
  <li>Minor improvements: qemud signal handling (Mark), don't shutdown or reboot
    domain0 (Kazuki Mizushima), QEmu version autodetection (Daniel Berrange),
    network UUIDs (Mark), speed up UUID domain lookups (Tatsuro Enokura and
    Daniel Berrange), support for paused QEmu CPU (Daniel Berrange), keymap
    VNC attribute support (Takahashi Tomohiro and Daniel Berrange), maximum
    number of virtual CPU (Masayuki Sunou), virtsh --readonly option (Rich
    Jones), python bindings for new functions (Daniel Berrange)</li>
  <li>Documentation updates especially on the XML formats</li>
</ul>

<h3>0.2.0: Feb 14 2007</h3>
<ul>
  <li>Various internal cleanups (Mark McLoughlin, Richard Jones,
      Daniel Berrange, Karel Zak)</li>
  <li>Bug fixes: avoid a crash in connect (Daniel Berrange), virsh args
      parsing (Richard Jones)</li>
  <li>Add support for QEmu and KVM virtualization (Daniel Berrange)</li>
  <li>Add support for network configuration (Mark McLoughlin)</li>
  <li>Minor improvements: regression testing (Daniel Berrange), 
      localization string updates</li>
</ul>

<h3>0.1.11: Jan 22 2007</h3>
<ul>
  <li>Finish XML &lt;-&gt; XM config files support</li>
  <li>Remove memory leak when freeing virConf objects</li>
  <li>Finishing inactive domain support (Daniel Berrange)</li>
  <li>Added a Relax-NG schemas to check XML instances</li>
</ul>

<h3>0.1.10: Dec 20 2006</h3>
<ul>
  <li>more localizations</li>
  <li>bug fixes: VCPU info breakages on xen 3.0.3, xenDaemonListDomains buffer overflow (Daniel Berrange), reference count bug when creating Xen domains (Daniel Berrange).</li>
  <li>improvements: support graphic framebuffer for Xen paravirt (Daniel Berrange), VNC listen IP range support (Daniel Berrange), support for default Xen config files and inactive domains of 3.0.4 (Daniel Berrange).</li>
</ul>

<h3>0.1.9: Nov 29 2006</h3>
<ul>
  <li>python bindings: release interpeter lock when calling C (Daniel Berrange)</li>
  <li>don't raise HTTP error when looking informations for a domain</li>
  <li>some refactoring to use the driver for all entry points</li>
  <li>better error reporting (Daniel Berrange)</li>
  <li>fix OS reporting when running as non-root</li>
  <li>provide XML parsing errors</li>
  <li>extension of the test framework (Daniel Berrange)</li>
  <li>fix the reconnect regression test</li>
  <li>python bindings: Domain instances now link to the Connect to avoid garbage collection and disconnect</li>
  <li>separate the notion of maximum memory and current use at the XML level</li>
  <li>Fix a memory leak (Daniel Berrange)</li>
  <li>add support for shareable drives</li>
  <li>add support for non-bridge style networking configs for guests(Daniel Berrange)</li>
  <li>python bindings: fix unsigned long marshalling (Daniel Berrange)</li>
  <li>new config APIs virConfNew() and virConfSetValue() to build configs from scratch</li>
  <li>hot plug device support based on Michel Ponceau patch</li>
  <li>added support for inactive domains, new APIs, various associated cleanup (Daniel Berrange)</li>
  <li>special device model for HVM guests (Daniel Berrange)</li>
  <li>add API to dump core of domains (but requires a patched xend)</li>
  <li>pygrub bootloader informations take over &lt;os&gt; informations</li>
  <li>updated the localization strings</li>
</ul>
<h3>0.1.8: Oct 16 2006</h3>
<ul>
  <li> Bug for system with page size != 4k</li>
  <li> vcpu number initialization (Philippe Berthault)</li>
  <li> don't label crashed domains as shut off (Peter Vetere)</li>
  <li> fix virsh man page (Noriko Mizumoto)</li>
  <li> blktapdd support for alternate drivers like blktap (Daniel Berrange)</li>
  <li> memory leak fixes (xend interface and XML parsing) (Daniel Berrange)</li>
  <li> compile fix</li>
  <li> mlock/munlock size fixes (Daniel Berrange)</li>
  <li> improve error reporting</li>
</ul>
<h3>0.1.7: Sep 29 2006</h3>
<ul>
  <li> fix a memory bug on getting vcpu informations from xend (Daniel Berrange)</li>
  <li> fix another problem in the hypercalls change in Xen changeset
       86d26e6ec89b when getting domain informations (Daniel Berrange)</li>
</ul>
<h3>0.1.6: Sep 22 2006</h3>
<ul>
  <li>Support for localization of strings using gettext (Daniel Berrange)</li>
  <li>Support for new Xen-3.0.3 cdrom and disk configuration (Daniel Berrange)</li>
  <li>Support for setting VNC port when creating domains with new
      xend config files (Daniel Berrange) </li>
  <li>Fix bug when running against xen-3.0.2 hypercalls (Jim Fehlig)</li>
  <li>Fix reconnection problem when talking directly to http xend</li>
</ul>
<h3>0.1.5: Sep 5 2006</h3>
<ul>
  <li>Support for new hypercalls change in Xen changeset 86d26e6ec89b</li>
  <li>bug fixes: virParseUUID() was wrong, netwoking for paravirt guestsi
      (Daniel Berrange), virsh on non-existent domains (Daniel Berrange),
      string cast bug when handling error in python (Pete Vetere), HTTP
      500 xend error code handling (Pete Vetere and Daniel Berrange)</li>
  <li>improvements: test suite for SEXPR &lt;-&gt; XML format conversions (Daniel
      Berrange), virsh output regression suite (Daniel Berrange), new environ
      variable VIRSH_DEFAULT_CONNECT_URI for the default URI when connecting
      (Daniel Berrange), graphical console support for paravirt guests
      (Jeremy Katz), parsing of simple Xen config files (with Daniel Berrange),
      early work on defined (not running) domains (Daniel Berrange),
      virsh output improvement (Daniel Berrange</li>
</ul>

<h3>0.1.4: Aug 16 2006</h3>
<ul>
  <li>bug fixes: spec file fix (Mark McLoughlin), error report problem (with
    Hugh Brock), long integer in Python bindings (with Daniel Berrange), XML
    generation bug for CDRom (Daniel Berrange), bug whem using number() XPath
    function (Mark McLoughlin), fix python detection code, remove duplicate
    initialization errors (Daniel Berrange)</li>
  <li>improvements: UUID in XML description (Peter Vetere), proxy code
    cleanup, virtual CPU and affinity support + virsh support (Michel
    Ponceau, Philippe Berthault, Daniel Berrange), port and tty informations
    for console in XML (Daniel Berrange), added XML dump to driver and proxy
    support (Daniel Berrange), extention of boot options with support for
    floppy and cdrom (Daniel Berrange), features block in XML to report/ask
    PAE, ACPI, APIC for HVM domains (Daniel Berrange), fail saide-effect
    operations when using read-only connection, large improvements to test
    driver (Daniel Berrange) </li>
  <li>documentation: spelling (Daniel Berrange), test driver examples.</li>
</ul>

<h3>0.1.3: Jul 11 2006</h3>
<ul>
  <li>bugfixes: build as non-root, fix xend access when root, handling of
    empty XML elements (Mark McLoughlin), XML serialization and parsing fixes
    (Mark McLoughlin), allow to create domains without disk (Mark
  McLoughlin),</li>
  <li>improvement: xenDaemonLookupByID from O(n^2) to O(n) (Daniel Berrange),
    support for fully virtualized guest (Jim Fehlig, DV, Mark McLoughlin)</li>
  <li>documentation: augmented to cover hvm domains</li>
</ul>

<h3>0.1.2: Jul 3 2006</h3>
<ul>
  <li>headers include paths fixup</li>
  <li>proxy mechanism for unpriviledged read-only access by httpu</li>
</ul>

<h3>0.1.1: Jun 21 2006</h3>
<ul>
  <li>building fixes: ncurses fallback (Jim Fehlig), VPATH builds (Daniel P.
    Berrange)</li>
  <li>driver cleanups: new entry points, cleanup of libvirt.c (with Daniel P.
    Berrange)</li>
  <li>Cope with API change introduced in Xen changeset 10277</li>
  <li>new test driver for regression checks (Daniel P. Berrange)</li>
  <li>improvements: added UUID to XML serialization, buffer usage (Karel
    Zak), --connect argument to virsh (Daniel P. Berrange),</li>
  <li>bug fixes: uninitialized memory access in error reporting, S-Expr
    parsing (Jim Fehlig, Jeremy Katz), virConnectOpen bug, remove a TODO in
    xs_internal.c</li>
  <li>documentation: Python examples (David Lutterkort), new Perl binding
    URL, man page update (Karel Zak)</li>
</ul>

<h3>0.1.0: Apr 10 2006</h3>
<ul>
  <li>building fixes: --with-xen-distdir option (Ronald Aigner), out of tree
    build and pkginfo cflag fix (Daniel Berrange)</li>
  <li>enhancement and fixes of the XML description format (David Lutterkort
    and Jim Fehlig)</li>
  <li>new APIs: for Node information and Reboot</li>
  <li>internal code cleanup: refactoring internals into a driver model, more
    error handling, structure sharing, thread safety and ref counting</li>
  <li>bug fixes: error message (Jim Meyering), error allocation in virsh (Jim
    Meyering), virDomainLookupByID (Jim Fehlig),</li>
  <li>documentation: updates on architecture, and format, typo fix (Jim
    Meyering)</li>
  <li>bindings: exception handling in examples (Jim Meyering), perl ones out
    of tree (Daniel Berrange)</li>
  <li>virsh: more options, create, nodeinfo (Karel Zak), renaming of some
    options (Karel Zak), use stderr only for errors (Karel Zak), man page
    (Andrew Puch)</li>
</ul>

<h3>0.0.6: Feb 28 2006</h3>
<ul>
  <li>add UUID lookup and extract API</li>
  <li>add error handling APIs both synchronous and asynchronous</li>
  <li>added minimal hook for error handling at the python level, improved the
    python bindings</li>
  <li>augment the documentation and tests to cover error handling</li>
</ul>

<h3>0.0.5: Feb 23 2006</h3>
<ul>
  <li>Added XML description parsing, dependance to libxml2, implemented the
    creation API virDomainCreateLinux()</li>
  <li>new APIs to lookup and name domain by UUID</li>
  <li>fixed the XML dump when using the Xend access</li>
  <li>Fixed a few more problem related to the name change</li>
  <li>Adding regression tests in python and examples in C</li>
  <li>web site improvement, extended the documentation to cover the XML
    format and Python API</li>
  <li>Added devhelp help for Gnome/Gtk programmers</li>
</ul>

<h3>0.0.4: Feb 10 2006</h3>
<ul>
  <li>Fix various bugs introduced in the name change</li>
</ul>

<h3>0.0.3: Feb 9 2006</h3>
<ul>
  <li>Switch name from from 'libvir' to libvirt</li>
  <li>Starting infrastructure to add code examples</li>
  <li>Update of python bindings for completeness</li>
</ul>

<h3>0.0.2: Jan 29 2006</h3>
<ul>
  <li>Update of the documentation, web site redesign (Diana Fong)</li>
  <li>integration of HTTP xend RPC based on libxend by Anthony Liquori for
    most operations</li>
  <li>Adding Save and Restore APIs</li>
  <li>extended the virsh command line tool (Karel Zak)</li>
  <li>remove xenstore transactions (Anthony Liguori)</li>
  <li>fix the Python bindings bug when domain and connections where freed</li>
</ul>

<h3>0.0.1: Dec 19 2005</h3>
<ul>
  <li>First release</li>
  <li>Basic management of existing Xen domains</li>
  <li>Minimal autogenerated Python bindings</li>
</ul>

<h2><a name="Introducti">Introduction</a></h2>

<p>Libvirt is a C toolkit to interact with the virtualization capabilities of
recent versions of Linux (and other OSes), but libvirt won't try to provide
all possible interfaces for interacting with the virtualization features.</p>

<p>To avoid ambiguity about the terms used here here are the definitions for
some of the specific concepts used in libvirt documentation:</p>
<ul>
  <li>a <strong>node</strong> is a single physical machine</li>
  <li>an <strong>hypervisor</strong> is a layer of software allowing to
    virtualize a node in a set of virtual machines with possibly different
    configurations than the node itself</li>
  <li>a <strong>domain</strong> is an instance of an operating system running
    on a virtualized machine provided by the hypervisor</li>
</ul>

<p style="text-align: center"><img
alt="Hypervisor and domains running on a node" src="node.gif"></p>

<p>Now we can define the goal of libvirt: to provide the lowest possible
generic and stable layer to manage domains on a node.</p>

<p>This implies the following:</p>
<ul>
  <li>the API should not be targetted to a single virtualization environment
    though Xen is the current default, which also means that some very
    specific capabilities which are not generic enough may not be provided as
    libvirt APIs</li>
  <li>the API should allow to do efficiently and cleanly all the operations
    needed to manage domains on a node</li>
  <li>the API will not try to provide hight level multi-nodes management
    features like load balancing, though they could be implemented on top of
    libvirt</li>
  <li>stability of the API is a big concern, libvirt should isolate
    applications from the frequent changes expected at the lower level of the
    virtualization framework</li>
</ul>

<p>So libvirt should be a building block for higher level management tools
and for applications focusing on virtualization of a single node (the only
exception being domain migration between node capabilities which may need to
be added at the libvirt level). Where possible libvirt should be extendable
to be able to provide the same API for remote nodes, however this is not the
case at the moment, the code currently handle only local node accesses
(extension for remote access support is being worked on, see <a
href="bugs.html">the mailing list</a> discussions about it).</p>

<h2><a name="architecture">libvirt architecture</a></h2>

<p>Currently libvirt supports 2 kind of virtualization, and its
internal structure is based on a driver model which simplifies adding new
engines:</p>

<ul>
  <li><a href="#Xen">Xen hypervisor</a></li>
  <li><a href="#QEmu">QEmu and KVM based virtualization</a></li>
  <li><a href="#drivers">the driver architecture</a></li>
</ul>

<h3><a name="Xen">Libvirt Xen support</a></h3>

<p>When running in a Xen environment, programs using libvirt have to execute
in "Domain 0", which is the primary Linux OS loaded on the machine. That OS
kernel provides most if not all of the actual drivers used by the set of
domains. It also runs the Xen Store, a database of informations shared by the
hypervisor, the kernels, the drivers and the xen daemon. Xend. The xen daemon
supervise the control and execution of the sets of domains. The hypervisor,
drivers, kernels and daemons communicate though a shared system bus
implemented in the hypervisor. The figure below tries to provide a view of
this environment:</p>
<img src="architecture.gif" alt="The Xen architecture">

<p>The library can be initialized in 2 ways depending on the level of
priviledge of the embedding program. If it runs with root access,
virConnectOpen() can be used, it will use three different ways to connect to
the Xen infrastructure:</p>
<ul>
  <li>a connection to the Xen Daemon though an HTTP RPC layer</li>
  <li>a read/write connection to the Xen Store</li>
  <li>use Xen Hypervisor calls</li>
  <li>when used as non-root libvirt connect to a proxy daemon running
      as root and providing read-only support</li>
</ul>

<p>The library will usually interact with the Xen daemon for any operation
changing the state of the system, but for performance and accuracy reasons
may talk directly to the hypervisor when gathering state informations at
least when possible (i.e. when the running program using libvirt has root
priviledge access).</p>

<p>If it runs without root access virConnectOpenReadOnly() should be used to
connect to initialize the library. It will then fork a libvirt_proxy
program running as root and providing read_only access to the API, this is
then only useful for reporting and monitoring.</p>

<h3><a name="QEmu">Libvirt QEmu and KVM support</a></h3>

<p>The model for QEmu and KVM is completely similar, basically KVM is based
on QEmu for the process controlling a new domain, only small details differs
between the two. In both case the libvirt API is provided by a controlling
process forked by libvirt in the background and which launch and control the
QEmu or KVM process. That program called libvirt_qemud talks though a specific
protocol to the library, and connects to the console of the QEmu process in
order to control and report on its status. Libvirt tries to expose all the
emulations models of QEmu, the selection is done when creating the new
domain, by specifying the architecture and machine type targetted.</p>

<p>The code controlling the QEmu process is available in the
<code>qemud/</code> directory.</p>

<h3><a name="drivers">the driver based architecture</a></h3>

<p>As the previous section explains, libvirt can communicate using different
channels with the current hypervisor, and should also be able to use
different kind of hypervisor. To simplify the internal design, code, ease
maintainance and simplify the support of other virtualization engine the
internals have been structured as one core component, the libvirt.c module
acting as a front-end for the library API and a set of hypvisor drivers
defining a common set of routines. That way the Xen Daemon accces, the Xen
Store one, the Hypervisor hypercall are all isolated in separate C modules
implementing at least a subset of the common operations defined by the
drivers present in driver.h:</p>
<ul>
  <li>xend_internal: implements the driver functions though the Xen
  Daemon</li>
  <li>xs_internal: implements the subset of the driver availble though the
    Xen Store</li>
  <li>xen_internal: provide the implementation of the functions possible via
    direct hypervisor access</li>
  <li>proxy_internal: provide read-only Xen access via a proxy, the proxy code
    is in the <code>proxy/</code>directory.</li>
  <li>xm_internal: provide support for Xen defined but not running
    domains.</li>
  <li>qemu_internal: implement the driver functions for QEmu and
    KVM virtualization engines. It also uses a qemud/ specific daemon
    which interracts with the QEmu process to implement libvirt API.</li>
  <li>test: this is a test driver useful for regression tests of the
    front-end part of libvirt.</li>
</ul>

<p>Note that a given driver may only implement a subset of those functions,
(for example saving a Xen domain state to disk and restoring it is only
possible though the Xen Daemon), in that case the driver entry points for
unsupported functions are initialized to NULL.</p>

<p></p>

<h2><a name="Downloads">Downloads</a></h2>

<p>The latest versions of libvirt can be found on the  <a
href="ftp://libvirt.org/libvirt/">libvirt.org</a> server ( <a
href="http://libvirt.org/sources/">HTTP</a>, <a
href="ftp://libvirt.org/libvirt/">FTP</a>). You will find there the released
versions as well as <a
href="http://libvirt.org/sources/libvirt-cvs-snapshot.tar.gz">snapshot
tarballs</a> updated from CVS head every hour</p>

<p>Anonymous <a href="http://ximbiot.com/cvs/cvshome/docs/">CVS</a> is also
available, first register onto the server:</p>

<p><code>cvs -d :pserver:anoncvs@libvirt.org:2401/data/cvs login</code></p>

<p>it will request a password, enter <strong>anoncvs</strong>. Then you can
checkout the development tree with:</p>

<p><code>cvs -d :pserver:anoncvs@libvirt.org:2401/data/cvs co
libvirt</code></p>

<p>Use ./autogen.sh to configure the local checkout, then <code>make</code>
and <code>make install</code>, as usual. All normal cvs commands are now
available except commiting to the base.</p>

<h2><a name="Format">XML Format</a></h2>

<p>This section describes the XML format used to represent domains, there are
variations on the format based on the kind of domains run and the options
used to launch them:</p>

<ul>
  <li><a href="#Normal1">Normal paravirtualized Xen domains</a></li>
  <li><a href="#Fully1">Fully virtualized Xen domains</a></li>
  <li><a href="#KVM1">KVM domains</a></li>
  <li><a href="#Net1">Networking options for QEmu and KVM</a></li>
  <li><a href="#QEmu1">QEmu domains</a></li>
  <li><a href="#Capa1">Discovering virtualization capabilities</a></li>
</ul>

<p>The formats try as much as possible to follow the same structure and reuse
elements and attributes where it makes sense.</p>

<h3 id="Normal"><a name="Normal1" id="Normal1">Normal paravirtualized Xen
guests</a>:</h3>

<p>The library use an XML format to describe domains, as input to <a
href="html/libvirt-libvirt.html#virDomainCreateLinux">virDomainCreateLinux()</a>
and as the output of <a
href="html/libvirt-libvirt.html#virDomainGetXMLDesc">virDomainGetXMLDesc()</a>,
the following is an example of the format as returned by the shell command
<code>virsh xmldump fc4</code> , where fc4 was one of the running domains:</p>
<pre>&lt;domain type='xen' <span style="color: #0071FF; background-color: #FFFFFF">id='18'</span>&gt;
  &lt;name&gt;fc4&lt;/name&gt;
  <span style="color: #00B200; background-color: #FFFFFF">&lt;os&gt;
    &lt;type&gt;linux&lt;/type&gt;
    &lt;kernel&gt;/boot/vmlinuz-2.6.15-1.43_FC5guest&lt;/kernel&gt;
    &lt;initrd&gt;/boot/initrd-2.6.15-1.43_FC5guest.img&lt;/initrd&gt;
    &lt;root&gt;/dev/sda1&lt;/root&gt;
    &lt;cmdline&gt; ro selinux=0 3&lt;/cmdline&gt;
  &lt;/os&gt;</span>
  &lt;memory&gt;131072&lt;/memory&gt;
  &lt;vcpu&gt;1&lt;/vcpu&gt;
  &lt;devices&gt;
    <span style="color: #FF0080; background-color: #FFFFFF">&lt;disk type='file'&gt;
      &lt;source file='/u/fc4.img'/&gt;
      &lt;target dev='sda1'/&gt;
    &lt;/disk&gt;</span>
    <span style="color: #0000FF; background-color: #FFFFFF">&lt;interface type='bridge'&gt;
      &lt;source bridge='xenbr0'/&gt;
      &lt;mac address='</span><span style="color: #0000FF; background-color: #FFFFFF"></span><span style="color: #0000FF; background-color: #FFFFFF">aa:00:00:00:00:11'/&gt;
      &lt;script path='/etc/xen/scripts/vif-bridge'/&gt;
    &lt;/interface&gt;</span>
    <span style="color: #FF8000; background-color: #FFFFFF">&lt;console tty='/dev/pts/5'/&gt;</span>
  &lt;/devices&gt;
&lt;/domain&gt;</pre>

<p>The root element must be called <code>domain</code> with no namespace, the
<code>type</code> attribute indicates the kind of hypervisor used, 'xen' is
the default value. The <code>id</code> attribute gives the domain id at
runtime (not however that this may change, for example if the domain is saved
to disk and restored). The domain has a few children whose order is not
significant:</p>
<ul>
  <li>name: the domain name, preferably ASCII based</li>
  <li>memory: the maximum memory allocated to the domain in kilobytes</li>
  <li>vcpu: the number of virtual cpu configured for the domain</li>
  <li>os: a block describing the Operating System, its content will be
    dependant on the OS type
    <ul>
      <li>type: indicate the OS type, always linux at this point</li>
      <li>kernel: path to the kernel on the Domain 0 filesystem</li>
      <li>initrd: an optional path for the init ramdisk on the Domain 0
        filesystem</li>
      <li>cmdline: optional command line to the kernel</li>
      <li>root: the root filesystem from the guest viewpoint, it may be
        passed as part of the cmdline content too</li>
    </ul>
  </li>
  <li>devices: a list of <code>disk</code>, <code>interface</code> and
    <code>console</code> descriptions in no special order</li>
</ul>

<p>The format of the devices and their type may grow over time, but the
following should be sufficient for basic use:</p>

<p>A <code>disk</code> device indicates a block device, it can have two
values for the type attribute either 'file' or 'block' corresponding to the 2
options availble at the Xen layer. It has two mandatory children, and one
optional one in no specific order:</p>
<ul>
  <li>source with a file attribute containing the path in Domain 0 to the
    file or a dev attribute if using a block device, containing the device
    name ('hda5' or '/dev/hda5')</li>
  <li>target indicates in a dev attribute the device where it is mapped in
    the guest</li>
  <li>readonly an optional empty element indicating the device is
  read-only</li>
</ul>

<p>An <code>interface</code> element describes a network device mapped on the
guest, it also has a type whose value is currently 'bridge', it also have a
number of children in no specific order:</p>
<ul>
  <li>source: indicating the bridge name</li>
  <li>mac: the optional mac address provided in the address attribute</li>
  <li>ip: the optional IP address provided in the address attribute</li>
  <li>script: the script used to bridge the interfcae in the Domain 0</li>
  <li>target: and optional target indicating the device name.</li>
</ul>

<p>A <code>console</code> element describes a serial console connection to
the guest. It has no children, and a single attribute <code>tty</code> which
provides the path to the Pseudo TTY on which the guest console can be
accessed</p>

<p>Life cycle actions for the domain can also be expressed in the XML format,
they drive what should be happening if the domain crashes, is rebooted or is
poweroff. There is various actions possible when this happen:</p>
<ul>
  <li>destroy: The domain is cleaned up (that's the default normal processing
    in Xen)</li>
  <li>restart: A new domain is started in place of the old one with the same
    configuration parameters</li>
  <li>preserve: The domain will remain in memory until it is destroyed
    manually, it won't be running but allows for post-mortem debugging</li>
  <li>rename-restart: a variant of the previous one but where the old domain
    is renamed before being saved to allow a restart</li>
</ul>

<p>The following could be used for a Xen production system:</p>
<pre>&lt;domain&gt;
  ...
  &lt;on_reboot&gt;restart&lt;/on_reboot&gt;
  &lt;on_poweroff&gt;destroy&lt;/on_poweroff&gt;
  &lt;on_crash&gt;rename-restart&lt;/on_crash&gt;
  ...
&lt;/domain&gt;</pre>

<p>While the format may be extended in various ways as support for more
hypervisor types and features are added, it is expected that this core subset
will remain functional in spite of the evolution of the library.</p>

<h3 id="Fully"><a name="Fully1" id="Fully1">Fully virtualized guests</a>
(added in 0.1.3):</h3>

<p>Here is an example of a domain description used to start a fully
virtualized (a.k.a. HVM) Xen domain. This requires hardware virtualization
support at the processor level but allows to run unmodified operating
systems:</p>
<pre>&lt;domain type='xen' id='3'&gt;
  &lt;name&gt;fv0&lt;/name&gt;
  &lt;uuid&gt;4dea22b31d52d8f32516782e98ab3fa0&lt;/uuid&gt;
  &lt;os&gt;
    <span style="color: #0000E5; background-color: #FFFFFF">&lt;type&gt;hvm&lt;/type&gt;</span>
    <span style="color: #0000E5; background-color: #FFFFFF">&lt;loader&gt;/usr/lib/xen/boot/hvmloader&lt;/loader&gt;</span>
    <span style="color: #0000E5; background-color: #FFFFFF">&lt;boot dev='hd'/&gt;</span>
  &lt;/os&gt;
  &lt;memory&gt;524288&lt;/memory&gt;
  &lt;vcpu&gt;1&lt;/vcpu&gt;
  &lt;on_poweroff&gt;destroy&lt;/on_poweroff&gt;
  &lt;on_reboot&gt;restart&lt;/on_reboot&gt;
  &lt;on_crash&gt;restart&lt;/on_crash&gt;
  &lt;features&gt;
     <span style="color: #E50000; background-color: #FFFFFF">&lt;pae/&gt;
     &lt;acpi/&gt;
     &lt;apic/&gt;</span>
  &lt;/features&gt;
  &lt;devices&gt;
    <span style="color: #0000E5; background-color: #FFFFFF">&lt;emulator&gt;/usr/lib/xen/bin/qemu-dm&lt;/emulator&gt;</span>
    &lt;interface type='bridge'&gt;
      &lt;source bridge='xenbr0'/&gt;
      &lt;mac address='00:16:3e:5d:c7:9e'/&gt;
      &lt;script path='vif-bridge'/&gt;
    &lt;/interface&gt;
    &lt;disk type='file'&gt;
      &lt;source file='/root/fv0'/&gt;
      &lt;target <span style="color: #0000E5; background-color: #FFFFFF">dev='hda'</span>/&gt;
    &lt;/disk&gt;
    &lt;disk type='file' <span style="color: #0000E5; background-color: #FFFFFF">device='cdrom'</span>&gt;
      &lt;source file='/root/fc5-x86_64-boot.iso'/&gt;
      &lt;target <span style="color: #0000E5; background-color: #FFFFFF">dev='hdc'</span>/&gt;
      &lt;readonly/&gt;
    &lt;/disk&gt;
    &lt;disk type='file' <span style="color: #0000E5; background-color: #FFFFFF">device='floppy'</span>&gt;
      &lt;source file='/root/fd.img'/&gt;
      &lt;target <span style="color: #0000E5; background-color: #FFFFFF">dev='fda'</span>/&gt;
    &lt;/disk&gt;
    <span style="color: #0000E5; background-color: #FFFFFF">&lt;graphics type='vnc' port='5904'/&gt;</span>
  &lt;/devices&gt;
&lt;/domain&gt;</pre>

<p>There is a few things to notice specifically for HVM domains:</p>
<ul>
  <li>the optional <code>&lt;features&gt;</code> block is used to enable
    certain guest CPU / system features. For HVM guests the following
    features are defined:
    <ul>
      <li><code>pae</code> - enable PAE memory addressing</li>
      <li><code>apic</code> - enable IO APIC</li>
      <li><code>acpi</code> - enable ACPI bios</li>
    </ul>
  </li>
  <li>the <code>&lt;os&gt;</code> block description is very different, first
    it indicates that the type is 'hvm' for hardware virtualization, then
    instead of a kernel, boot and command line arguments, it points to an os
    boot loader which will extract the boot informations from the boot device
    specified in a separate boot element. The <code>dev</code> attribute on
    the <code>boot</code> tag can be one of:
    <ul>
      <li><code>fd</code> - boot from first floppy device</li>
      <li><code>hd</code> - boot from first harddisk device</li>
      <li><code>cdrom</code> - boot from first cdrom device</li>
    </ul>
  </li>
  <li>the <code>&lt;devices&gt;</code> section includes an emulator entry
    pointing to an additional program in charge of emulating the devices</li>
  <li>the disk entry indicates in the dev target section that the emulation
    for the drive is the first IDE disk device hda. The list of device names
    supported is dependant on the Hypervisor, but for Xen it can be any IDE
    device <code>hda</code>-<code>hdd</code>, or a floppy device
    <code>fda</code>, <code>fdb</code>. The <code>&lt;disk&gt;</code> element
    also supports a 'device' attribute to indicate what kinda of hardware to
    emulate. The following values are supported:
    <ul>
      <li><code>floppy</code> - a floppy disk controller</li>
      <li><code>disk</code> - a generic hard drive (the default it
      omitted)</li>
      <li><code>cdrom</code> - a CDROM device</li>
    </ul>
    For Xen 3.0.2 and earlier a CDROM device can only be emulated on the
    <code>hdc</code> channel, while for 3.0.3 and later, it can be emulated
    on any IDE channel.</li>
  <li>the <code>&lt;devices&gt;</code> section also include at least one
    entry for the graphic device used to render the os. Currently there is
    just 2 types possible 'vnc' or 'sdl'. If the type is 'vnc', then an
    additional <code>port</code> attribute will be present indicating the TCP
    port on which the VNC server is accepting client connections.</li>
</ul>

<p>It is likely that the HVM description gets additional optional elements
and attributes as the support for fully virtualized domain expands,
especially for the variety of devices emulated and the graphic support
options offered.</p>

<h3><a name="KVM1">KVM domain (added in 0.2.0)</a></h3>

<p>Support for the <a href="http://kvm.qumranet.com/">KVM virtualization</a>
is provided in recent Linux kernels (2.6.20 and onward). This requires
specific hardware with acceleration support and the availability of the
special version of the <a
href="http://fabrice.bellard.free.fr/qemu/">QEmu</a> binary. Since this
relies on QEmu for the machine emulation like fully virtualized guests the
XML description is quite similar, here is a simple example:</p>
<pre>&lt;domain <span style="color: #FF0000; background-color: #FFFFFF">type='kvm'</span>&gt;
  &lt;name&gt;demo2&lt;/name&gt;
  &lt;uuid&gt;4dea24b3-1d52-d8f3-2516-782e98a23fa0&lt;/uuid&gt;
  &lt;memory&gt;131072&lt;/memory&gt;
  &lt;vcpu&gt;1&lt;/vcpu&gt;
  &lt;os&gt;
    &lt;type&gt;hvm&lt;/type&gt;
  &lt;/os&gt;
  &lt;devices&gt;
    <span style="color: #FF0000; background-color: #FFFFFF">&lt;emulator&gt;/home/user/usr/kvm-devel/bin/qemu-system-x86_64&lt;/emulator&gt;</span>
    &lt;disk type='file' device='disk'&gt;
      &lt;source file='/home/user/fedora/diskboot.img'/&gt;
      &lt;target dev='hda'/&gt;
    &lt;/disk&gt;
    &lt;interface <span style="color: #FF0000; background-color: #FFFFFF">type='user'</span>&gt;
      &lt;mac address='24:42:53:21:52:45'/&gt;
    &lt;/interface&gt;
    &lt;graphics type='vnc' port='-1'/&gt;
  &lt;/devices&gt;
&lt;/domain&gt;</pre>

<p>The specific points to note if using KVM are:</p>
<ul>
  <li>the top level domain element carries a type of 'kvm'</li>
  <li>the &lt;devices&gt; emulator points to the special qemu binary required
    for KVM</li>
  <li>networking interface definitions definitions are somewhat different due
    to a different model from Xen see below</li>
</ul>

<p>except those points the options should be quite similar to Xen HVM
ones.</p>

<h3><a name="Net1">Networking options for QEmu and KVM (added in 0.2.0)</a></h3>

<p>The networking support in the QEmu and KVM case is more flexible, and
support a variety of options:</p>
<ol>
  <li>Userspace SLIRP stack
    <p>Provides a virtual LAN with NAT to the outside world. The virtual
    network has DHCP &amp; DNS services and will give the guest VM addresses
    starting from <code>10.0.2.15</code>. The default router will be
    <code>10.0.2.2</code> and the DNS server will be <code>10.0.2.3</code>.
    This networking is the only option for unprivileged users who need their
    VMs to have outgoing access. Example configs are:</p>
    <pre>&lt;interface type='user'/&gt;</pre>
    <pre>
&lt;interface type='user'&gt;                                                  
  &lt;mac address="11:22:33:44:55:66:/&gt;                                     
&lt;/interface&gt;
    </pre>
  </li>
  <li>Virtual network
    <p>Provides a virtual network using a bridge device in the host.
    Depending on the virtual network configuration, the network may be
    totally isolated,NAT'ing to aan explicit network device, or NAT'ing to
    the default route. DHCP and DNS are provided on the virtual network in
    all cases and the IP range can be determined by examining the virtual
    network config with '<code>virsh net-dumpxml &lt;network
    name&gt;</code>'. There is one virtual network called'default' setup out
    of the box which does NAT'ing to the default route and has an IP range of
    <code>192.168.22.0/255.255.255.0</code>. Each guest will have an
    associated tun device created with a name of vnetN, which can also be
    overriden with the &lt;target&gt; element. Example configs are:</p>
    <pre>&lt;interface type='network'&gt;
  &lt;source network='default'/&gt;
&lt;/interface&gt;

&lt;interface type='network'&gt;
  &lt;source network='default'/&gt;
  &lt;target dev='vnet7'/&gt;
  &lt;mac address="11:22:33:44:55:66:/&gt;
&lt;/interface&gt;
    </pre>
  </li>
  <li>Bridge to to LAN
    <p>Provides a bridge from the VM directly onto the LAN. This assumes
    there is a bridge device on the host which has one or more of the hosts
    physical NICs enslaved. The guest VM will have an associated tun device
    created with a name of vnetN, which can also be overriden with the
    &lt;target&gt; element. The tun device will be enslaved to the bridge.
    The IP range / network configuration is whatever is used on the LAN. This
    provides the guest VM full incoming &amp; outgoing net access just like a
    physical machine. Examples include:</p>
    <pre>&lt;interface type='bridge'&gt;
 &lt;source dev='br0'/&gt;
&lt;/interface&gt;

&lt;interface type='bridge'&gt;
  &lt;source dev='br0'/&gt;
  &lt;target dev='vnet7'/&gt;
  &lt;mac address="11:22:33:44:55:66:/&gt;
&lt;/interface&gt;       &lt;interface type='bridge'&gt;
         &lt;source dev='br0'/&gt;
         &lt;target dev='vnet7'/&gt;
         &lt;mac address="11:22:33:44:55:66:/&gt;
       &lt;/interface&gt;</pre>
  </li>
  <li>Generic connection to LAN
    <p>Provides a means for the administrator to execute an arbitrary script
    to connect the guest's network to the LAN. The guest will have a tun
    device created with a name of vnetN, which can also be overriden with the
    &lt;target&gt; element. After creating the tun device a shell script will
    be run which is expected to do whatever host network integration is
    required. By default this script is called /etc/qemu-ifup but can be
    overriden.</p>
    <pre>&lt;interface type='ethernet'/&gt;

&lt;interface type='ethernet'&gt;
  &lt;target dev='vnet7'/&gt;
  &lt;script path='/etc/qemu-ifup-mynet'/&gt;
&lt;/interface&gt;</pre>
  </li>
  <li>Multicast tunnel
    <p>A multicast group is setup to represent a virtual network. Any VMs
    whose network devices are in the same multicast group can talk to each
    other even across hosts. This mode is also available to unprivileged
    users. There is no default DNS or DHCP support and no outgoing network
    access. To provide outgoing network access, one of the VMs should have a
    2nd NIC which is connected to one of the first 4 network types and do the
    appropriate routing. The multicast protocol is compatible with that used
    by user mode linux guests too. The source address used must be from the
    multicast address block.</p>
    <pre>&lt;interface type='mcast'&gt;
  &lt;source address='230.0.0.1' port='5558'/&gt;
&lt;/interface&gt;</pre>
  </li>
  <li>TCP tunnel
    <p>A TCP client/server architecture provides a virtual network. One VM
    provides the server end of the netowrk, all other VMS are configured as
    clients. All network traffic is routed between the VMs via the server.
    This mode is also available to unprivileged users. There is no default
    DNS or DHCP support and no outgoing network access. To provide outgoing
    network access, one of the VMs should have a 2nd NIC which is connected
    to one of the first 4 network types and do the appropriate routing.</p>
    <p>Example server config:</p>
    <pre>&lt;interface type='server'&gt;
  &lt;source address='192.168.0.1' port='5558'/&gt;
&lt;/interface&gt;</pre>
    <p>Example client config:</p>
    <pre>&lt;interface type='client'&gt;
  &lt;source address='192.168.0.1' port='5558'/&gt;
&lt;/interface&gt;</pre>
  </li>
</ol>

<p>To be noted, options 2, 3, 4 are also supported by Xen VMs, so it is
possible to use these configs to have networking with both Xen &amp;
QEMU/KVMs connected to each other.</p>

<h3>Q<a name="QEmu1">Emu domain (added in 0.2.0)</a></h3>

<p>Libvirt support for KVM and QEmu is the same code base with only minor
changes. The configuration is as a result nearly identical, the only changes
are related to QEmu ability to emulate <a
href="http://www.qemu.org/status.html">various CPU type and hardware
platforms</a>, and kqemu support (QEmu own kernel accelerator when the
emulated CPU is i686 as well as the target machine):</p>
<pre>&lt;domain <span style="color: #FF0000; background-color: #FFFFFF">type='qemu'</span>&gt;
  &lt;name&gt;QEmu-fedora-i686&lt;/name&gt;
  &lt;uuid&gt;c7a5fdbd-cdaf-9455-926a-d65c16db1809&lt;/uuid&gt;
  &lt;memory&gt;219200&lt;/memory&gt;
  &lt;currentMemory&gt;219200&lt;/currentMemory&gt;
  &lt;vcpu&gt;2&lt;/vcpu&gt;
  &lt;os&gt;
    <span style="color: #FF0000; background-color: #FFFFFF">&lt;type arch='i686' machine='pc'&gt;hvm&lt;/type&gt;</span>
    &lt;boot dev='cdrom'/&gt;
  &lt;/os&gt;
  &lt;devices&gt;
    <span style="color: #FF0000; background-color: #FFFFFF">&lt;emulator&gt;/usr/bin/qemu&lt;/emulator&gt;</span>
    &lt;disk type='file' device='cdrom'&gt;
      &lt;source file='/home/user/boot.iso'/&gt;
      &lt;target dev='hdc'/&gt;
      &lt;readonly/&gt;
    &lt;/disk&gt;
    &lt;disk type='file' device='disk'&gt;
      &lt;source file='/home/user/fedora.img'/&gt;
      &lt;target dev='hda'/&gt;
    &lt;/disk&gt;
    &lt;interface type='network'&gt;
      &lt;source name='default'/&gt;
    &lt;/interface&gt;
    &lt;graphics type='vnc' port='-1'/&gt;
  &lt;/devices&gt;
&lt;/domain&gt;</pre>

<p>The difference here are:</p>
<ul>
  <li>the value of type on top-level domain, it's 'qemu' or kqemu if asking
    for <a href="http://www.qemu.org/kqemu-tech.html">kernel assisted
    acceleration</a></li>
  <li>the os type block defines the architecture to be emulated, and
    optionally the machine type, see the discovery API below</li>
  <li>the emulator string must point to the right emulator for that
    architecture</li>
</ul>

<h3><a name="Capa1">Discovering virtualization capabilities (Added in 0.2.1)</a></h3>

<p>As new virtualization engine support gets added to libvirt, and to handle
cases like QEmu supporting a variety of emulations, a query interface has
been added in 0.2.1 allowing to list the set of supported virtualization
capabilities on the host:</p>
<pre>    char * virConnectGetCapabilities (virConnectPtr conn);</pre>

<p>The value returned is an XML document listing the virtualization
capabilities of the host and virtualization engine to which
<code>@conn</code> is connected. One can test it using <code>virsh</code>
command line tool command '<code>capabilities</code>', it dumps the XML
associated to the current connection. For example in the case of a 64 bits
machine with hardware virtualization capabilities enabled in the chip and
BIOS you will see</p>
<pre>&lt;capabilities&gt;
  <span style="color: #E50000; background-color: #FFFFFF">&lt;host&gt;
    &lt;cpu&gt;
      &lt;arch&gt;x86_64&lt;/arch&gt;
      &lt;features&gt;
        &lt;vmx/&gt;
      &lt;/features&gt;
    &lt;/cpu&gt;
  &lt;/host&gt;</span>

  &lt;!-- xen-3.0-x86_64 --&gt;
  <span style="color: #0000E5; background-color: #FFFFFF">&lt;guest&gt;
    &lt;os_type&gt;xen&lt;/os_type&gt;
    &lt;arch name="x86_64"&gt;
      &lt;wordsize&gt;64&lt;/wordsize&gt;
      &lt;domain type="xen"&gt;&lt;/domain&gt;
      &lt;emulator&gt;/usr/lib64/xen/bin/qemu-dm&lt;/emulator&gt;
    &lt;/arch&gt;
    &lt;features&gt;
    &lt;/features&gt;
  &lt;/guest&gt;</span>

  &lt;!-- hvm-3.0-x86_32 --&gt;
  <span style="color: #00B200; background-color: #FFFFFF">&lt;guest&gt;
    &lt;os_type&gt;hvm&lt;/os_type&gt;
    &lt;arch name="i686"&gt;
      &lt;wordsize&gt;32&lt;/wordsize&gt;
      &lt;domain type="xen"&gt;&lt;/domain&gt;
      &lt;emulator&gt;/usr/lib/xen/bin/qemu-dm&lt;/emulator&gt;
      &lt;machine&gt;pc&lt;/machine&gt;
      &lt;machine&gt;isapc&lt;/machine&gt;
      &lt;loader&gt;/usr/lib/xen/boot/hvmloader&lt;/loader&gt;
    &lt;/arch&gt;
    &lt;features&gt;
    &lt;/features&gt;
  &lt;/guest&gt;</span>
  ...
&lt;/capabilities&gt;</pre>

<p>The fist block (in red) indicates the host hardware capbilities, currently
it is limited to the CPU properties but other information may be available,
it shows the CPU architecture, and the features of the chip (the feature
block is similar to what you will find in a Xen fully virtualized domain
description).</p>

<p>The second block (in blue) indicates the paravirtualization support of the
Xen support, you will see the os_type of xen to indicate a paravirtual
kernel, then architecture informations and potential features.</p>

<p>The third block (in green) gives similar informations but when running a
32 bit OS fully virtualized with Xen using the hvm support.</p>

<p>This section is likely to be updated and augmented in the future, see <a
href="https://www.redhat.com/archives/libvir-list/2007-March/msg00215.html">the
discussion</a> which led to the capabilities format in the mailing-list
archives.</p>

<h2><a name="Python" id="Python">Binding for Python</a></h2>

<p>Libvirt comes with direct support for the Python language (just make sure
you installed the libvirt-python package if not compiling from sources). Also
note that Daniel Berrange provides <a
href="http://search.cpan.org/~danberr/Sys-Virt-0.1.0/">bindings for Perl</a>
too.</p>

<p>The Python binding should be complete and are mostly automatically
generated from the formal description of the API in xml. The bindings are
articulated around 2 classes <code>virConnect</code> and virDomain mapping to
the C types. Functions in the C API taking either type as argument then
becomes methods for the classes, their name is just stripped from the
virConnect or virDomain(Get) prefix and the first letter gets converted to
lower case, for example the C functions:</p>

<p><code>int <a
href="html/libvirt-libvirt.html#virConnectNumOfDomains">virConnectNumOfDomains</a>
(virConnectPtr conn);</code></p>

<p><code>int <a
href="html/libvirt-libvirt.html#virDomainSetMaxMemory">virDomainSetMaxMemory</a>
(virDomainPtr domain, unsigned long memory);</code></p>

<p>become</p>

<p><code>virConn::numOfDomains(self)</code></p>

<p><code>virDomain::setMaxMemory(self, memory)</code></p>

<p>This process is fully automated, you can get a summary of the conversion
in the file libvirtclass.txt present in the python dir or in the docs.There
is a couple of function who don't map directly to their C counterparts due to
specificities in their argument conversions:</p>
<ul>
  <li><code><a
    href="html/libvirt-libvirt.html#virConnectListDomains">virConnectListDomains</a></code>
    is replaced by <code>virDomain::listDomainsID(self)</code> which returns
    a list of the integer ID for the currently running domains</li>
  <li><code><a
    href="html/libvirt-libvirt.html#virDomainGetInfo">virDomainGetInfo</a></code>
    is replaced by <code>virDomain::info()</code> which returns a list of
    <ol>
      <li>state: one of the state values (virDomainState)</li>
      <li>maxMemory: the maximum memory used by the domain</li>
      <li>memory: the current amount of memory used by the domain</li>
      <li>nbVirtCPU: the number of virtual CPU</li>
      <li>cpuTime: the time used by the domain in nanoseconds</li>
    </ol>
  </li>
</ul>

<p>So let's look at a simple example inspired from the <code>basic.py</code>
test found in <code>python/tests/</code> in the source tree:</p>
<pre>import <span style="color: #0071FF; background-color: #FFFFFF">libvirt</span>
import sys

conn = <span style="color: #0071FF; background-color: #FFFFFF">libvirt</span>.openReadOnly(None)
if conn == None:
    print 'Failed to open connection to the hypervisor'
    sys.exit(1)

try:
    dom0 = conn.<span style="color: #007F00; background-color: #FFFFFF">lookupByName</span>("Domain-0")
except:
    print 'Failed to find the main domain'
    sys.exit(1)

print "Domain 0: id %d running %s" % (dom0.<span style="color: #FF0080; background-color: #FFFFFF">ID</span>(), dom0.<span style="color: #FF0080; background-color: #FFFFFF">OSType</span>())
print dom0.<span style="color: #FF0080; background-color: #FFFFFF">info</span>()</pre>

<p>There is not much to comment about it, it really is a straight mapping
from the C API, the only points to notice are:</p>
<ul>
  <li>the import of the module called <code><span
    style="color: #0071FF; background-color: #FFFFFF">libvirt</span></code></li>
  <li>getting a connection to the hypervisor, in that case using the
    openReadOnly function allows the code to execute as a normal user.</li>
  <li>getting an object representing the Domain 0 using <span
    style="color: #007F00; background-color: #FFFFFF">lookupByName</span></li>
  <li>if the domain is not found a libvirtError exception will be raised</li>
  <li>extracting and printing some informations about the domain using
    various <span
    style="color: #E50073; background-color: #FFFFFF">methods</span>
    associated to the virDomain class.</li>
</ul>

<h2><a name="Errors" id="Errors">Handling of errors</a></h2>

<p>The main goals of libvirt when it comes to error handling are:</p>
<ul>
  <li>provide as much detail as possible</li>
  <li>provide the informations as soon as possible</li>
  <li>dont force the library user into one style of error handling</li>
</ul>

<p>As result the library provide both synchronous, callback based and
asynchronous error reporting. When an error happens in the library code the
error is logged, allowing to retrieve it later and if the user registered an
error callback it will be called synchronously. Once the call to libvirt ends
the error can be detected by the return value and the full information for
the last logged error can be retrieved.</p>

<p>To avoid as much as prossible troubles with a global variable in a
multithreaded environment, libvirt will associate when possible the errors to
the current connection they are related to, that way the error is stored in a
dynamic structure which can be made thread specific. Error callback can be
set specifically to a connection with</p>

<p>So error handling in the code is the following:</p>
<ol>
  <li>if the error can be associated to a connection for example when failing
    to look up a domain
    <ol>
      <li>if there is a callback associated to the connection set with <a
        href="html/libvirt-virterror.html#virConnSetErrorFunc">virConnSetErrorFunc</a>,
        call it with the error informations</li>
      <li>otherwise if there is a global callback set with <a
        href="html/libvirt-virterror.html#virSetErrorFunc">virSetErrorFunc</a>,
        call it with the error information</li>
      <li>otherwise call <a
        href="html/libvirt-virterror.html#virDefaultErrorFunc">virDefaultErrorFunc</a>
        which is the default error function of the library issuing the error
        on stderr</li>
      <li>save the error in the connection for later retrieval with <a
        href="html/libvirt-virterror.html#virConnGetLastError">virConnGetLastError</a></li>
    </ol>
  </li>
  <li>otherwise like when failing to create an hypervisor connection:
    <ol>
      <li>if there is a global callback set with <a
        href="html/libvirt-virterror.html#virSetErrorFunc">virSetErrorFunc</a>,
        call it with the error information</li>
      <li>otherwise call <a
        href="html/libvirt-virterror.html#virDefaultErrorFunc">virDefaultErrorFunc</a>
        which is the default error function of the library issuing the error
        on stderr</li>
      <li>save the error in the connection for later retrieval with <a
        href="html/libvirt-virterror.html#virGetLastError">virGetLastError</a></li>
    </ol>
  </li>
</ol>

<p>In all cases the error informations are provided as a <a
href="html/libvirt-virterror.html#virErrorPtr">virErrorPtr</a> pointer to
read-only structure <a
href="html/libvirt-virterror.html#virError">virError</a> containing the
following fields:</p>
<ul>
  <li>code: an error number from the <a
    href="html/libvirt-virterror.html#virErrorNumber">virErrorNumber</a>
  enum</li>
  <li>domain: an enum indicating which part of libvirt raised the error see
    <a
  href="html/libvirt-virterror.html#virErrorDomain">virErrorDomain</a></li>
  <li>level: the error level, usually VIR_ERR_ERROR, though there is room for
    warnings like VIR_ERR_WARNING</li>
  <li>message: the full human-readable formatted string of the error</li>
  <li>conn: if available a pointer to the <a
    href="html/libvirt-libvirt.html#virConnectPtr">virConnectPtr</a>
    connection to the hypervisor where this happened</li>
  <li>dom: if available a pointer to the <a
    href="html/libvirt-libvirt.html#virDomainPtr">virDomainPtr</a> domain
    targetted in the operation</li>
</ul>

<p>and then extra raw informations about the error which may be initialized
to 0 or NULL if unused</p>
<ul>
  <li>str1, str2, str3: string informations, usually str1 is the error
    message format</li>
  <li>int1, int2: integer informations</li>
</ul>

<p>So usually, setting up specific error handling with libvirt consist of
registering an handler with with <a
href="html/libvirt-virterror.html#virSetErrorFunc">virSetErrorFunc</a> or
with <a
href="html/libvirt-virterror.html#virConnSetErrorFunc">virConnSetErrorFunc</a>,
chech the value of the code value, take appropriate action, if needed let
libvirt print the error on stderr by calling <a
href="html/libvirt-virterror.html#virDefaultErrorFunc">virDefaultErrorFunc</a>.
For asynchronous error handing, set such a function doing nothing to avoid
the error being reported on stderr, and call virConnGetLastError or
virGetLastError when an API call returned an error value. It can be a good
idea to use <a
href="html/libvirt-virterror.html#virResetLastError">virResetError</a> or <a
href="html/libvirt-virterror.html#virConnResetLastError">virConnResetLastError</a>
once an error has been processed fully.</p>

<p>At the python level, there only a global reporting callback function at
this point, see the error.py example about it:</p>
<pre>def handler(ctxt, err):
    global errno

    #print "handler(%s, %s)" % (ctxt, err)
    errno = err

libvirt.registerErrorHandler(handler, 'context') </pre>

<p>the second argument to the registerErrorHandler function is passed as the
fist argument of the callback like in the C version. The error is a tuple
containing the same field as a virError in C, but cast to Python.</p>

<h2><a name="FAQ" id="FAQ">FAQ</a></h2>

<p>Table of Contents:</p>
<ul>
  <li><a href="FAQ.html#License">License(s)</a></li>
  <li><a href="FAQ.html#Installati">Installation</a></li>
  <li><a href="FAQ.html#Compilatio">Compilation</a></li>
  <li><a href="FAQ.html#Developer">Developer corner</a></li>
</ul>

<h3><a name="License">License</a>(s)</h3>
<ol>
  <li><em>Licensing Terms for libvirt</em>
    <p>libvirt is released under the <a
    href="http://www.opensource.org/licenses/lgpl-license.html">GNU Lesser
    General Public License</a>, see the file COPYING.LIB in the distribution
    for the precise wording. The only library that libvirt depends upon is
    the Xen store access library which is also licenced under the LGPL.</p>
  </li>
  <li><em>Can I embed libvirt in a proprietary application ?</em>
    <p>Yes. The LGPL allows you to embed libvirt into a proprietary
    application. It would be graceful to send-back bug fixes and improvements
    as patches for possible incorporation in the main development tree. It
    will decrease your maintainance costs anyway if you do so.</p>
  </li>
</ol>

<h3><a name="Installati">Installation</a></h3>
<ol>
  <li><em>Where can I get libvirt</em> ?
    <p>The original distribution comes from <a
    href="ftp://libvirt.org/libvirt/">ftp://libvirt.org/libvirt/</a>.</p>
  </li>
  <li><em>I can't install the libvirt/libvirt-devel RPM packages due to
    failed dependencies</em>
    <p>The most generic solution is to re-fetch the latest src.rpm , and
    rebuild it locally with</p>
    <p><code>rpm --rebuild libvirt-xxx.src.rpm</code>.</p>
    <p>If everything goes well it will generate two binary rpm packages (one
    providing the shared libs and virsh, and the other one, the -devel
    package, providing includes, static libraries and scripts needed to build
    applications with libvirt that you can install locally.</p>
    <p>One can also rebuild the RPMs from a tarball:</p>
    <p><code>rpmbuild -ta libdir-xxx.tar.gz</code></p>
    <p>Or from a configured tree with:</p>
    <p><code>make rpm</code></p>
  </li>
  <li><em>Failure to use the API for non-root users</em>
    <p>Large parts of the API may only be accessible with root priviledges,
    however the read only access to the xenstore data doesnot have to be
    forbidden to user, at least for monitoring purposes. If "virsh dominfo"
    fails to run as an user, change the mode of the xenstore read-only socket
    with:</p>
    <p><code>chmod 666 /var/run/xenstored/socket_ro</code></p>
    <p>and also make sure that the Xen Daemon is running correctly with local
    HTTP server enabled, this is defined in
    <code>/etc/xen/xend-config.sxp</code> which need the following line to be
    enabled:</p>
    <p><code>(xend-http-server yes)</code></p>
    <p>If needed restart the xend daemon after making the change with the
    following command run as root:</p>
    <p><code>service xend restart</code></p>
  </li>
</ol>

<h3><a name="Compilatio">Compilation</a></h3>
<ol>
  <li><em>What is the process to compile libvirt ?</em>
    <p>As most UNIX libraries libvirt follows the "standard":</p>
    <p><code>gunzip -c libvirt-xxx.tar.gz | tar xvf -</code></p>
    <p><code>cd libvirt-xxxx</code></p>
    <p><code>./configure --help</code></p>
    <p>to see the options, then the compilation/installation proper</p>
    <p><code>./configure [possible options]</code></p>
    <p><code>make</code></p>
    <p><code>make install</code></p>
    <p>At that point you may have to rerun ldconfig or a similar utility to
    update your list of installed shared libs.</p>
  </li>
  <li><em>What other libraries are needed to compile/install libvirt ?</em>
    <p>Libvirt requires libxenstore, which is usually provided by the xen
    packages as well as the public headers to compile against libxenstore.</p>
  </li>
  <li><em>I use the CVS version and there is no configure script</em>
    <p>The configure script (and other Makefiles) are generated. Use the
    autogen.sh script to regenerate the configure script and Makefiles,
    like:</p>
    <p><code>./autogen.sh --prefix=/usr --disable-shared</code></p>
  </li>
</ol>

<h3><a name="Developer">Developer</a> corner</h3>
<ol>
  <li><em>Troubles compiling or linking programs using libvirt</em>
    <p>To simplify the process of reusing the library, libvirt comes with
    pkgconfig support, which can be used directly from autoconf support or
    via the pkg-config command line tool, like:</p>
    <p><code>pkg-config libvirt --libs</code></p>
  </li>
</ol>

<h2><a name="Reporting">Reporting bugs and getting help</a></h2>

<p>There is a mailing-list <a
href="mailto:libvir-list@redhat.com">libvir-list@redhat.com</a> for libvirt,
with an  <a href="https://www.redhat.com/archives/libvir-list/">on-line
archive</a>. Please subscribe to this list before posting by visiting the <a
href="https://www.redhat.com/mailman/listinfo/libvir-list">associated Web</a>
page and follow the instructions. Patches with explanations and provided as
attachments are really appreciated and will be discussed on the mailing list.
If possible generate the patches by using cvs diff -u in a CVS checkout.</p>

<p>We use Red Hat Bugzilla to track bugs to libvirt. If you want to report a
bug, please check <a href="http://bugzilla.redhat.com/bugzilla/buglist.cgi?component=libvirt&amp;component=libvirt-devel&amp;component=libvirt-python&amp;bug_status=ASSIGNED&amp;bug_status=INVESTIGATE&amp;bug_status=NEW&amp;bug_status=REOPENED&amp;bug_status=VERIFIED&amp;short_desc_type=allwordssubstr&amp;short_desc=&amp;long_desc_type=allwordssubstr&amp;long_desc=&amp;Search=Search">the existing open bugs</a>, then if yours isn't a duplicate of
an existing bug, <a href="http://bugzilla.redhat.com/bugzilla/enter_bug.cgi?product=Fedora%20Core&amp;component=libvirt">log a new bug</a>. It may be good
to post to the <a href="mailto:libvir-list@redhat.com">mailing-list</a>
too if the issue looks serious, thanks !</p>

<h2><a name="Remote">Remote support</a></h2>

<p>
<b>NB. Remote support is available only as a <a
href="https://www.redhat.com/archives/libvir-list/">series of
patches posted on libvir-list</a> against <a
href="http://libvirt.org/downloads.html">libvirt CVS</a>.  It is only
for experimental use at the moment.</b>
&mdash; Richard Jones, 2007-04-18.
</p>

<p>
Libvirt allows you to access hypervisors running on remote
machines through authenticated and encrypted connections.
</p>

<h3><a name="Remote_basic_usage">Basic usage</a></h3>

<p>
On the remote machine, <code>libvirtd</code> should be running.
See <a href="#Remote_libvirtd_configuration">the section
on configuring libvirtd</a> for more information.
</p>

<p>
To tell libvirt that you want to access a remote resource,
you should supply a hostname in the normal URI that is passed
to <code>virConnectOpen</code> (or <code>virsh -c ...</code>).
For example, if you normally use <code>qemu:///system</code>
to access the system-wide QEMU daemon, then to access
the system-wide QEMU daemon on a remote machine called
<code>oirase</code> you would use <code>qemu://oirase/system</code>.
</p>

<p>
The <a href="#Remote_URI_reference">section on remote URIs</a>
describes in more detail these remote URIs.
</p>

<p>
From an API point of view, apart from the change in URI, the
API should behave the same.  For example, ordinary calls
are routed over the remote connection transparently, and
values or errors from the remote side are returned to you
as if they happened locally.  Some differences you may notice:
</p>

<ul>
<li> Additional errors can be generated, specifically ones
relating to failures in the remote transport itself. </li>
<li> Remote calls are handled synchronously, so they will be
much slower than, say, direct hypervisor calls. </li>
</ul>

<h3><a name="Remote_transports">Transports</a></h3>

<p>
Remote libvirt supports a range of transports:
</p>

<dl>
<dt> tls </dt>
<dd> <a href="http://en.wikipedia.org/wiki/Transport_Layer_Security"
   title="Transport Layer Security">TLS</a>
 1.0 (SSL 3.1) authenticated and encrypted TCP/IP socket, usually
 listening on a public port number.  To use this you will need to
 <a href="#Remote_certificates"
 title="Generating TLS certificates">generate client and
 server certificates</a>.
 The standard port is 16514.
 </dd>

<dt> unix </dt>
<dd> Unix domain socket.  Since this is only accessible on the
 local machine, it is not encrypted, and uses Unix permissions or
 SELinux for authentication.
 The standard socket names are
 <code>/var/run/libvirt/libvirt-sock</code> and
 <code>/var/run/libvirt/libvirt-sock-ro</code> (the latter
 for read-only connections).
 </dd>

<dt> ssh </dt>
<dd> Transported over an ordinary
 <a href="http://www.openssh.com/" title="OpenSSH homepage">ssh
 (secure shell)</a> connection.
 Requires <a href="http://netcat.sourceforge.net/">Netcat (nc)</a>
 installed on the remote machine, and the remote libvirtd should
 be listening on the unix transport.  You should use some sort of
 ssh key management (eg.
 <a href="http://mah.everybody.org/docs/ssh"
 title="Using ssh-agent with ssh">ssh-agent</a>)
 otherwise programs which use
 this transport will stop to ask for a password. </dd>

<dt> ext </dt>
<dd> Any external program which can make a connection to the
 remote machine by means outside the scope of libvirt. </dd>

<dt> tcp </dt>
<dd> Unencrypted TCP/IP socket.  Not recommended for production
 use, this is normally disabled, but an administrator can enable
 it for testing or use over a trusted network.
 The standard port is 16509.
 </dd>
</dl>

<p>
The default transport, if no other is specified, is <code>tls</code>.
</p>

<h3><a name="Remote_URI_reference">Remote URIs</a></h3>

<p>
Remote URIs have the general form ("[...]" meaning an optional part):
</p>

<p>
<code>driver</code>[<code>+transport</code>]<code>://</code>[<code>username@</code>]<code>hostname</code>[<code>:port</code>]<code>/</code>[<code>path</code>][<code>?extraparameters</code>]
</p>

<p>
Either the transport or the hostname must be given in order
to distinguish this from a local URI.
</p>

<p>
Some examples:
</p>

<ul>
<li> <code>xen+ssh://rjones@towada/</code> <br/> &mdash; Connect to a
remote Xen hypervisor on host <code>towada</code> using ssh transport and ssh
username <code>rjones</code>.
</li>

<li> <code>xen://towada/</code> <br/> &mdash; Connect to a
remote Xen hypervisor on host <code>towada</code> using TLS.
</li>

<li> <code>xen://towada/?no_verify=1</code> <br/> &mdash; Connect to a
remote Xen hypervisor on host <code>towada</code> using TLS.  Do not verify
the server's certificate.
</li>

<li> <code>qemu+unix:///system?socket=/opt/libvirt/run/libvirt/libvirt-sock</code> <br/> &mdash;
Connect to the local qemu instances over a non-standard
Unix socket (the full path to the Unix socket is
supplied explicitly in this case).
</li>

<li> <code>test+tcp://localhost:5000/default</code> <br/> &mdash;
Connect to a libvirtd daemon offering unencrypted TCP/IP connections
on localhost port 5000 and use the test driver with default
settings.
</li>

</ul>

<h4><a name="Remote_URI_parameters">Extra parameters</a></h4>

<p>
Extra parameters can be added to remote URIs as part
of the query string (the part following <q><code>?</code></q>).
Remote URIs understand the extra parameters shown below.
Any others are passed unmodified through to the back end.
Note that parameter values must be
<a href="http://xmlsoft.org/html/libxml-uri.html#xmlURIEscapeStr">URI-escaped</a>.
</p>

<table>
<tr>
<th> Name </th>
<th> Transports </th>
<th> Meaning </th>
</tr>

<tr>
<td> <code>name</code> </td>
<td> <i>any transport</i> </td>
<td>
  The name passed to the remote virConnectOpen function.  The
  name is normally formed by removing transport, hostname, port
  number, username and extra parameters from the remote URI, but in certain
  very complex cases it may be better to supply the name explicitly.
</td>
</tr>
<tr> <td colspan="2"></td>
<td> Example: <code>name=qemu:///system</code> </td>
</tr>

<tr>
<td> <code>command</code> </td>
<td> ssh, ext </td>
<td>
  The external command.  For ext transport this is required.
  For ssh the default is <code>ssh</code>.
  The PATH is searched for the command.
</td>
</tr>
<tr> <td colspan="2"></td>
<td> Example: <code>command=/opt/openssh/bin/ssh</code> </td>
</tr>

<tr>
<td> <code>socket</code> </td>
<td> unix, ssh </td>
<td>
  The path to the Unix domain socket, which overrides the
  compiled-in default.  For ssh transport, this is passed to
  the remote netcat command (see next).
</td>
</tr>
<tr> <td colspan="2"></td>
<td> Example: <code>socket=/opt/libvirt/run/libvirt/libvirt-sock</code> </td>
</tr>

<tr>
<td> <code>netcat</code> </td>
<td> ssh </td>
<td>
  The name of the netcat command on the remote machine.
  The default is <code>nc</code>.  For ssh transport, libvirt
  constructs an ssh command which looks like:

<pre>
<i>command</i> -p <i>port</i> [-l <i>username</i>] <i>hostname</i> <i>netcat</i> -U <i>socket</i>
</pre>

  where <i>port</i>, <i>username</i>, <i>hostname</i> can be
  specified as part of the remote URI, and <i>command</i>, <i>netcat</i>
  and <i>socket</i> come from extra parameters (or
  sensible defaults).

</td>
</tr>
<tr> <td colspan="2"></td>
<td> Example: <code>netcat=/opt/netcat/bin/nc</code> </td>
</tr>

<tr>
<td> <code>no_verify</code> </td>
<td> tls </td>
<td>
  If set to a non-zero value, this disables client checks of the
  server's certificate.  Note that to disable server checks of
  the client's certificate or IP address you must
  <a href="#Remote_libvirtd_configuration">change the libvirtd
  configuration</a>.
</td>
</tr>
<tr> <td colspan="2"></td>
<td> Example: <code>no_verify=1</code> </td>
</tr>

</table>

<h3><a name="Remote_certificates">Generating TLS certificates</a></h3>

<p>
<i>This section to follow.</i>
</p>


<h3><a name="Remote_libvirtd_configuration">libvirtd configuration</a></h3>

<p>
<i>This section to follow.</i>
</p>


<h3><a name="Remote_IPv6">IPv6 support</a></h3>

<p>
IPv6 has received some limited testing and should work.  Problems with
libvirt and IPv6 should be reported as <a href="bugs.html">bugs</a>.
</p>

<h3><a name="Remote_limitations">Limitations</a></h3>

<ul>
<li> Remote storage: To be fully useful, particularly for
creating new domains, it should be possible to enumerate
and provision storage on the remote machine.  This is currently
in the design phase. </li>

<li> Migration: We expect libvirt will support migration,
and obviously remote support is what makes migration worthwhile.
This is also in the design phase.  Issues <a
href="https://www.redhat.com/mailman/listinfo/libvir-list"
title="libvir-list mailing list">to discuss</a> include
which path the migration data should follow (eg. client to
client direct, or client to server to client) and security.
</li>

<li> Fine-grained authentication: libvirt in general,
but in particular the remote case should support more
fine-grained authentication for operations, rather than
just read-write/read-only as at present.
</li>
</ul>

<p>
Please come and discuss these issues and more on <a
href="https://www.redhat.com/mailman/listinfo/libvir-list"
title="libvir-list mailing list">the mailing list</a>.
</p>

<h3><a name="Remote_implementation_notes">Implementation notes</a></h3>

<p>
The current implementation uses <a
href="http://en.wikipedia.org/wiki/External_Data_Representation"
title="External Data Representation">XDR</a>-encoded packets with a
simple remote procedure call implementation which also supports
asynchronous messaging and asynchronous and out-of-order replies,
although these latter features are not used at the moment.
</p>

<p>
The implementation should be considered <b>strictly internal</b> to
libvirt and <b>subject to change at any time without notice</b>.  If
you wish to talk to libvirtd, link to libvirt.  If there is a problem
that means you think you need to use the protocol directly, please
first discuss this on <a
href="https://www.redhat.com/mailman/listinfo/libvir-list"
title="libvir-list mailing list">the mailing list</a>.
</p>

<p>
The messaging protocol is described in
<code>qemud/remote_protocol.x</code>.
</p>

<p>
Authentication and encryption (for TLS) is done using <a
href="http://www.gnu.org/software/gnutls/" title="GnuTLS project
page">GnuTLS</a> and the RPC protocol is unaware of this layer.
</p>

<p>
Protocol messages are sent using a simple 32 bit length word (encoded
XDR int) followed by the message header (XDR
<code>remote_message_header</code>) followed by the message body.  The
length count includes the length word itself, and is measured in
bytes.  Maximum message size is <code>REMOTE_MESSAGE_MAX</code> and to
avoid denial of services attacks on the XDR decoders strings are
individually limited to <code>REMOTE_STRING_MAX</code> bytes.  In the
TLS case, messages may be split over TLS records, but a TLS record
cannot contain parts of more than one message.  In the common RPC case
a single <code>REMOTE_CALL</code> message is sent from client to
server, and the server then replies synchronously with a single
<code>REMOTE_REPLY</code> message, but other forms of messaging are
also possible.
</p>

<p>
The protocol contains support for multiple program types and protocol
versioning, modelled after SunRPC.
</p>

</body>
</html>