[libvirt] [PATCH] segfailt in openvz driver
by Evgeniy Sokolov
There was segmentation failt in openvz driver during 'create' command
if domain with same id exist.
Segfault was because of using structure (print warning with name of
container) which is free.
other:
use openvzError instead of openvzLog.
16 years, 3 months
[libvirt] Any outstanding patches?
by Richard W.M. Jones
While Daniel Veillard is away, if you have any outstanding patches
which have been approved but not committed, please reply to this
message with a pointer to the mailing list message or patch.
Rich.
--
Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones
virt-df lists disk usage of guests without needing to install any
software inside the virtual machine. Supports Linux and Windows.
http://et.redhat.com/~rjones/virt-df/
16 years, 3 months
[libvirt] [PATCH] Wrong comments
by Anton Protopopov
Hi!
In comments to structure 'virDomainInfo' there was reference to smth of type
'virDomainFlags'. Actually, the type of this field must be 'virDomainState'.
Patch includes another one such replacement.
There is another place where 'virDomainFlags' mentioned:
src/xend_internal.c. I think, that argument must be of type
'virConnectFlags',
but it is marked as ATTRIBUTE_UNUSED, so I leave the comment as it is :)
And why are you writing
unsigned char state;
instead of
virDomainState state;
?
Anton
16 years, 3 months
[libvirt] [PATCH] spec file enhancements for minimal "client-only" build
by Ben Guthro
Attached is a patch to add some additional options to the spec file,
allowing for more flexibility when building.
With this patch, it allows you to build a "client-only" version of
libvirt for machines that may have an interest in communicating over the
remote driver, but do not have a hypervisor themselves.
It enables command line overrides of control variables for the build
eg:
rpmbuild $(RPM_FLAGS) --nodeps \
--without xen \
--without xen_proxy \
--without qemu \
--without openvz \
--without lxc \
--without polkit \
--without libvirtd \
--without avahi \
--without sasl \
--without python \
--without devel \
-ta --clean libvirt*.tar.gz
Signed-off-by: Ben Guthro <bguthro(a)virtualiron.com>
diff --git a/libvirt.spec.in b/libvirt.spec.in
index 18b2127..666fb83 100644
--- a/libvirt.spec.in
+++ b/libvirt.spec.in
@@ -1,11 +1,16 @@
# -*- rpm-spec -*-
-%define with_xen 1
-%define with_xen_proxy 1
-%define with_qemu 1
-%define with_openvz 1
-%define with_lxc 1
-%define with_polkit 0
+%define with_xen 0%{!?_without_xen:1}
+%define with_xen_proxy 0%{!?_without_xen_proxy:1}
+%define with_qemu 0%{!?_without_qemu:1}
+%define with_openvz 0%{!?_without_openvz:1}
+%define with_lxc 0%{!?_without_lxc:1}
+%define with_sasl 0%{!?_without_sasl:1}
+%define with_avahi 0%{!?_without_avahi:1}
+%define with_polkit 0%{!?_without_polkit:0}
+%define with_python 0%{!?_without_python:1}
+%define with_libvirtd 0%{!?_without_libvirtd:1}
+%define with_devel 0%{!?_without_devel:1}
# Xen is available only on i386 x86_64 ia64
%ifnarch i386 i686 x86_64 ia64
@@ -23,7 +28,7 @@
%endif
%if 0%{fedora} >= 8
-%define with_polkit 1
+%define with_polkit 0%{!?_without_polkit:1}
%define with_xen_proxy 0
%endif
@@ -46,10 +51,12 @@ Requires: iptables
# So remote clients can access libvirt over SSH tunnel
# (client invokes 'nc' against the UNIX socket on the server)
Requires: nc
+%if %{with_sasl}
Requires: cyrus-sasl
# Not technically required, but makes 'out-of-box' config
# work correctly & doesn't have onerous dependencies
Requires: cyrus-sasl-md5
+%endif
%if %{with_polkit}
Requires: PolicyKit >= 0.6
%endif
@@ -82,14 +89,18 @@ BuildRequires: readline-devel
BuildRequires: ncurses-devel
BuildRequires: gettext
BuildRequires: gnutls-devel
+%if %{with_avahi}
BuildRequires: avahi-devel
+%endif
BuildRequires: libselinux-devel
BuildRequires: dnsmasq
BuildRequires: bridge-utils
%if %{with_qemu}
BuildRequires: qemu
%endif
+%if %{with_sasl}
BuildRequires: cyrus-sasl-devel
+%endif
%if %{with_polkit}
BuildRequires: PolicyKit-devel >= 0.6
%endif
@@ -119,6 +130,7 @@ BuildRequires: gawk
Libvirt is a C toolkit to interact with the virtualization capabilities
of recent versions of Linux (and other OSes).
+%if %{with_devel}
%package devel
Summary: Libraries, includes, etc. to compile with the libvirt library
Group: Development/Libraries
@@ -132,7 +144,9 @@ Obsoletes: libvir-devel
%description devel
Includes and documentations for the C library providing an API to use
the virtualization capabilities of recent versions of Linux (and other
OSes).
+%endif
+%if %{with_python}
%package python
Summary: Python bindings for the libvirt library
Group: Development/Libraries
@@ -144,6 +158,7 @@ The libvirt-python package contains a module that
permits applications
written in the Python programming language to use the interface
supplied by the libvirt library to use the virtualization capabilities
of recent versions of Linux (and other OSes).
+%endif
%prep
%setup -q
@@ -190,6 +205,7 @@ install -d -m 0755
$RPM_BUILD_ROOT%{_localstatedir}/lib/libvirt/images/
# Default dir for kernel+initrd images defnied in SELinux policy
install -d -m 0755 $RPM_BUILD_ROOT%{_localstatedir}/lib/libvirt/boot/
+%if %{with_qemu}
# We don't want to install /etc/libvirt/qemu/networks in the main
%files list
# because if the admin wants to delete the default network completely,
we don't
# want to end up re-incarnating it on every RPM upgrade.
@@ -200,14 +216,33 @@ rm -f
$RPM_BUILD_ROOT%{_sysconfdir}/libvirt/qemu/networks/default.xml
rm -f
$RPM_BUILD_ROOT%{_sysconfdir}/libvirt/qemu/networks/autostart/default.xml
# Strip auto-generated UUID - we need it generated per-install
sed -i -e "/<uuid>/d"
$RPM_BUILD_ROOT%{_datadir}/libvirt/networks/default.xml
+%endif
%find_lang %{name}
+# If we are not building with certain options, we need to remove files
+# so that we don't get Installed (but unpackaged) errors
+%if ! %{with_devel}
+rm -rf $RPM_BUILD_ROOT%{_datadir}/gtk-doc/html/libvirt
+rm -rf $RPM_BUILD_ROOT%{_includedir}/libvirt
+rm -rf $RPM_BUILD_ROOT%{_libdir}/lib*.so
+rm -rf $RPM_BUILD_ROOT%{_libdir}/pkgconfig/libvirt.pc
+%endif
+
+%if ! %{with_python}
+rm -rf $RPM_BUILD_ROOT%{_datadir}/doc/libvirt-python-%{version}
+%endif
+
+%if ! %{with_qemu}
+rm -rf $RPM_BUILD_ROOT%{_sysconfdir}/libvirt/qemu.conf
+%endif
+
%clean
rm -fr %{buildroot}
%post
/sbin/ldconfig
+%if %{with_libvirtd}
# We want to install the default network for initial RPM installs
# or on the first upgrade from a non-network aware libvirt only.
# We check this by looking to see if the daemon is already installed
@@ -222,12 +257,15 @@ then
fi
/sbin/chkconfig --add libvirtd
+%endif
%preun
+%if %{with_libvirtd}
if [ $1 = 0 ]; then
/sbin/service libvirtd stop 1>/dev/null 2>&1
/sbin/chkconfig --del libvirtd
fi
+%endif
%postun
/sbin/ldconfig
@@ -240,40 +278,73 @@ fi
%{_bindir}/virsh
%{_libdir}/lib*.so.*
%dir %attr(0700, root, root) %{_sysconfdir}/libvirt/
+
+%if %{with_qemu}
%dir %attr(0700, root, root) %{_sysconfdir}/libvirt/qemu/
%dir %attr(0700, root, root) %{_sysconfdir}/libvirt/qemu/networks/
%dir %attr(0700, root, root) %{_sysconfdir}/libvirt/qemu/networks/autostart
+%endif
+
+%if %{with_libvirtd}
%{_sysconfdir}/rc.d/init.d/libvirtd
%config(noreplace) %{_sysconfdir}/sysconfig/libvirtd
%config(noreplace) %{_sysconfdir}/libvirt/libvirtd.conf
+%endif
+
+%if %{with_qemu}
%config(noreplace) %{_sysconfdir}/libvirt/qemu.conf
+%endif
+
+%if %{with_sasl}
%config(noreplace) %{_sysconfdir}/sasl2/libvirt.conf
+%endif
+
+%if %{with_qemu}
%dir %{_datadir}/libvirt/
%dir %{_datadir}/libvirt/networks/
%{_datadir}/libvirt/networks/default.xml
+%endif
+
%dir %{_localstatedir}/run/libvirt/
%dir %{_localstatedir}/lib/libvirt/
%dir %attr(0700, root, root) %{_localstatedir}/lib/libvirt/images/
%dir %attr(0700, root, root) %{_localstatedir}/lib/libvirt/boot/
-%{_datadir}/augeas/lenses/libvirtd.aug
+
+%if %{with_qemu}
%{_datadir}/augeas/lenses/libvirtd_qemu.aug
-%{_datadir}/augeas/lenses/tests/test_libvirtd.aug
%{_datadir}/augeas/lenses/tests/test_libvirtd_qemu.aug
+%endif
+
+%if %{with_libvirtd}
+%{_datadir}/augeas/lenses/libvirtd.aug
+%{_datadir}/augeas/lenses/tests/test_libvirtd.aug
+%endif
+
%if %{with_polkit}
%{_datadir}/PolicyKit/policy/org.libvirt.unix.policy
%endif
+
+%if %{with_qemu}
%dir %attr(0700, root, root) %{_localstatedir}/log/libvirt/qemu/
+%endif
+
%if %{with_xen_proxy}
%attr(4755, root, root) %{_libexecdir}/libvirt_proxy
%endif
-%attr(0755, root, root) %{_libexecdir}/libvirt_parthelper
+
%if %{with_lxc}
%attr(0755, root, root) %{_libexecdir}/libvirt_lxc
%endif
+
+%if %{with_libvirtd}
+%attr(0755, root, root) %{_libexecdir}/libvirt_parthelper
%attr(0755, root, root) %{_sbindir}/libvirtd
+%endif
+
%doc docs/*.rng
%doc docs/*.xml
+%if %{with_devel}
%files devel
%defattr(-, root, root)
@@ -289,7 +360,9 @@ fi
%doc docs/*.html docs/html docs/*.gif
%doc docs/examples
%doc docs/libvirt-api.xml
+%endif
+%if %{with_python}
%files python
%defattr(-, root, root)
@@ -300,6 +373,7 @@ fi
%doc python/TODO
%doc python/libvirtclass.txt
%doc docs/examples/python
+%endif
%changelog
* Mon Sep 8 2008 Daniel Veillard <veillard(a)redhat.com> - 0.4.5-1
16 years, 3 months
[libvirt] [PATCH] kvm: maxVCPU runtime detection
by Guido Günther
Hi,
with recent linux kernels we can detect the maximum number of virtual
cpus at runtime via an ioctl. Possible patch attached. It does this on
every call to qemudGetMaxVCPUs. Would you prefer something that does
this only once in qemudStartup()?
Cheers,
-- Guido
16 years, 3 months
[libvirt] Calling libvirt from C# (Mono/.Net)
by Richard W.M. Jones
As an experiment I managed to call libvirt from C# (using Mono, but
similar code should work for .Net). This email should give interested
people enough information to produce a proper set of bindings for
libvirt.
----------------------------------------------------------------------
using System;
using System.Runtime.InteropServices;
[StructLayout (LayoutKind.Sequential)]
class NodeInfo {
[MarshalAs (UnmanagedType.ByValTStr, SizeConst=32)]
public string model;
public ulong memory;
public uint cpus;
public uint mhz;
public uint nodes;
public uint sockets;
public uint cores;
public uint threads;
}
class Connection {
IntPtr _handle;
[DllImport ("libvirt.so")]
static extern IntPtr virConnectOpen ([In] string name);
public Connection (string name) {
_handle = virConnectOpen (name);
}
[DllImport ("libvirt.so")]
static extern IntPtr virConnectClose (IntPtr h);
~Connection () {
virConnectClose (_handle);
}
[DllImport ("libvirt.so")]
static extern int virNodeGetInfo (IntPtr h, [Out] NodeInfo info);
public NodeInfo nodeGetInfo () {
NodeInfo info = new NodeInfo ();
if (virNodeGetInfo (_handle, info) == -1) {
// throw new LibvirtException ("virNodeGetInfo");
}
return info;
}
}
class Test
{
public static void Main ()
{
Connection conn = new Connection ("test:///default");
NodeInfo info = conn.nodeGetInfo ();
Console.WriteLine ("Results of virNodeGetInfo:");
Console.WriteLine ("model = {0}", info.model);
Console.WriteLine ("memory = {0}", info.memory);
Console.WriteLine ("cpus = {0}", info.cpus);
Console.WriteLine ("mhz = {0}", info.mhz);
Console.WriteLine ("nodes = {0}", info.nodes);
Console.WriteLine ("sockets = {0}", info.sockets);
Console.WriteLine ("cores = {0}", info.cores);
Console.WriteLine ("threads = {0}", info.threads);
}
}
----------------------------------------------------------------------
Compile and run this code with:
$ gmcs libvirt-test.c#
$ LIBVIRT_DEBUG=1 mono ./libvirt-test.exe
----------------------------------------------------------------------
DEBUG: libvirt.c: virInitialize (register drivers)
DEBUG: libvirt.c: virConnectOpen (name=test:///default)
DEBUG: libvirt.c: do_open (name "test:///default" to URI components:
scheme test
opaque (null)
authority (null)
server (null)
user (null)
port 0
path /default
)
DEBUG: libvirt.c: do_open (trying driver 0 (Test) ...)
DEBUG: libvirt.c: do_open (driver 0 Test returned SUCCESS)
DEBUG: libvirt.c: do_open (network driver 0 Test returned SUCCESS)
DEBUG: libvirt.c: do_open (storage driver 0 Test returned SUCCESS)
DEBUG: libvirt.c: virNodeGetInfo (conn=0xa8d560, info=0x7fff9530c070)
Results of virNodeGetInfo:
model = i686
memory = 3145728
cpus = 16
mhz = 1400
nodes = 2
sockets = 2
cores = 2
threads = 2
DEBUG: libvirt.c: virConnectClose (conn=0xa8d560)
DEBUG: hash.c: virUnrefConnect (unref connection 0xa8d560 test:///default 1)
DEBUG: hash.c: virReleaseConnect (release connection 0xa8d560 test:///default)
----------------------------------------------------------------------
Rich.
--
Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones
virt-top is 'top' for virtual machines. Tiny program with many
powerful monitoring features, net stats, disk stats, logging, etc.
http://et.redhat.com/~rjones/virt-top
16 years, 3 months
[libvirt] network domain name patch
by JJ Reynolds
Hi all,
This is a followup to the excellent patch which allows static IP address assignment. The problem with that patch is that you can only set the host name and cannot set a FQDN because dnsmasq will, as a security measure, not allow it unless --domain is specified.
This patch adds support for adding <domain name="my.domain" /> to the network config file. With that stanza, one can then use FQDNs on the static host assignments, and this should be the domain reported for any clients that request it. If <domain name .. ./> is not specified in the config file, then there is no change in behaviour.
As a special case, you can also set the domain name to "#", whereupon dnsmasq interprets that to use the domain of the host OS.
Example default.xml:
<network>
<name>default</name>
<uuid>0098abb7-ff94-4df9-aa78-e4c3fe636a3d</uuid>
<bridge name="virbr0" />
<domain name="mynet.net" />
<forward/>
<ip address="192.168.122.1" netmask="255.255.255.0">
<dhcp>
<host name="vm1.mynet.net" mac="00:16:3e:24:a5:84" ip="192.168.122.101" />
<host name='vm2.mynet.net' mac="00:16:3e:1f:9a:95" ip="192.168.122.102" />
<range start="192.168.122.2" end="192.168.122.99" />
</dhcp>
</ip>
</network>
Here is the patch:
diff -r -U 3 libvirt-0.4.4/src/network_conf.c libvirt-0.4.4jjr/src/network_conf.c
--- libvirt-0.4.4/src/network_conf.c 2008-08-20 09:20:01.000000000 -0400
+++ libvirt-0.4.4jjr/src/network_conf.c 2008-09-06 14:03:56.000000000 -0400
@@ -326,6 +326,9 @@
VIR_FREE(tmp);
}
+ /* Parse network domain information */
+ def->domain = virXPathString(conn, "string(./domain[1]/@name)", ctxt);
+
/* Parse bridge information */
def->bridge = virXPathString(conn, "string(./bridge[1]/@name)", ctxt);
tmp = virXPathString(conn, "string(./bridge[1]/@stp)", ctxt);
diff -r -U 3 libvirt-0.4.4/src/network_conf.h libvirt-0.4.4jjr/src/network_conf.h
--- libvirt-0.4.4/src/network_conf.h 2008-08-20 09:20:01.000000000 -0400
+++ libvirt-0.4.4jjr/src/network_conf.h 2008-09-06 14:03:56.000000000 -0400
@@ -57,6 +57,7 @@
char *name;
char *bridge; /* Name of bridge device */
+ char *domain;
unsigned long delay; /* Bridge forward delay (ms) */
int stp : 1; /* Spanning tree protocol */
diff -r -U 3 libvirt-0.4.4/src/qemu_driver.c libvirt-0.4.4jjr/src/qemu_driver.c
--- libvirt-0.4.4/src/qemu_driver.c 2008-08-29 03:20:02.000000000 -0400
+++ libvirt-0.4.4jjr/src/qemu_driver.c 2008-09-06 14:15:28.000000000 -0400
@@ -1105,6 +1105,7 @@
1 + /* --keep-in-foreground */
1 + /* --strict-order */
1 + /* --bind-interfaces */
+ (network->def->domain?2:0) + /* --domain name */
2 + /* --pid-file "" */
2 + /* --conf-file "" */
/*2 + *//* --interface virbr0 */
@@ -1136,6 +1137,11 @@
APPEND_ARG(*argv, i++, "--strict-order");
APPEND_ARG(*argv, i++, "--bind-interfaces");
+ if (network->def->domain) {
+ APPEND_ARG(*argv, i++, "--domain");
+ APPEND_ARG(*argv, i++, network->def->domain);
+ }
+
APPEND_ARG(*argv, i++, "--pid-file");
APPEND_ARG(*argv, i++, "");
Thanks!
-JJ Reynolds
16 years, 3 months
[libvirt] LXC: making the private root filesystem more secure
by Daniel P. Berrange
When I wrote the private root filesystem stuff for LXC (which I just
committed) I noted that we couldn't actually make this secure, because
someone inside the chroot can just 'mknod' and access the host devices.
What I completely forgot was that cgroups as of 2.6.26 has device ACLs
If we place every container in a cgroup (which was planned anyway), then
we can trivially prevent containers accessing host devices
One time setup
mount -t cgroups /dev/cgroups
mkdir /dev/cgroups/libvirt
mkdir /dev/cgroups/libvirt/lxc
For each new container 'NAME'
mkdir /dev/cgroups/libvirt/lxc/{NAME}
echo "a" > /dev/cgroups/libvirt/lxc/{NAME}/devices.deny
echo "c 1:3 rwm" > /dev/cgroups/libvirt/lxc/{NAME}/devices.allow
echo "c 1:5 rwm" > /dev/cgroups/libvirt/lxc/{NAME}/devices.allow
echo "c 1:7 rwm" > /dev/cgroups/libvirt/lxc/{NAME}/devices.allow
echo "c 5:1 rwm" > /dev/cgroups/libvirt/lxc/{NAME}/devices.allow
echo "c 1:8 rwm" > /dev/cgroups/libvirt/lxc/{NAME}/devices.allow
echo "c 1:9 rwm" > /dev/cgroups/libvirt/lxc/{NAME}/devices.allow
This denies all devices, and then allows null, zero, full, console, random
and urandom. Allowing use of 'random' is debatable.
The 'devpts' namespace stuff is also needed to provide private PTYs.
The 'user' namespace stuff is needed to prevent an unprivileged user
in the host OS from killing off processes with same UID inside the
container. There looks to be active patchsets for both of these being
discussed, so we're getting close to having a genuinely useful
container based virt driver with LXC
Daniel
--
|: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :|
|: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :|
|: http://autobuild.org -o- http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|
16 years, 3 months
[libvirt] virDomainGetInfo() returns 0 although the domain is not running anymore
by Johannes Formann
Hello,
if I understand the API-documentation, virDomainGetInfo() should return
-1 when the domain does not exists anymore, or some other problem
happens.
But with this code below
...
if (virDomainGetInfo(domainPtr,domainInfo) == 0 ) {
if (debugLevel > 15) {puts("virDomainGetInfo() returns 0");}
if (domainInfo->state == VIR_DOMAIN_NOSTATE ||
domainInfo->state == VIR_DOMAIN_RUNNING ||
domainInfo->state == VIR_DOMAIN_BLOCKED ||
domainInfo->state == VIR_DOMAIN_PAUSED ) {
return(false);
}
}
...
I notized, virDomainGetInfo() always returns zero, even if libvirt
prints an error.
If I destroy a domain (xm destroy) I'll get this Message with each call
of virDomainGetInfo():
"libvir: Xen Daemon error : GET operation failed:"
But virDomainGetInfo() still returns zero, a bug, or did I
missunderstood the API?
regards
Johannes
16 years, 3 months
