- Users - libvirt List Archives

[libvirt-users] [PATCHv11 0/6] libvirt/qemu - persistent modification of devices
by KAMEZAWA Hiroyuki 22 Apr '11

22 Apr '11

Here is v11. Fixed comments/bugs and updated against the latest libvirt.git. Changes v10->v11: - fixed comments on each patches - fixed cgroup handling in patch 3. - fixed MODIFY_CURRENT handling in patch 4. most of diff comes from refactoring qemu/qemu_driver.c -- conf/domain_conf.c | 40 ++ conf/domain_conf.h | 5 libvirt_private.syms | 3 qemu/qemu_driver.c | 727 +++++++++++++++++++++++++++++++-------------------- 4 files changed, 500 insertions(+), 275 deletions(-) Thanks, -Kame

2 9

[libvirt-users] No IOMMU found. Unable to assign device
by Marco van Beek 21 Apr '11

21 Apr '11

Hi, I am playing around with PCI passthrough and came across some posts that said you could not do PCI passthrough unless you had IOMMU hardware. which it would appear I don't Is this the case? Am I flogging a dead horse? Thanks, Marco -- Marco van Beek ========================================== Supporting Role Ltd. Grove Park Studios, 188-192 Sutton Court Rd, London, W4 3HR ========================================== T: 0870 757 2824 F: 0870 757 2826 M: 0788 770 3604 E: mvanbeek(a)supporting-role.co.uk W: http://www.supporting-role.co.uk/ ========================================== Company Registered in England No. 03658568 ==========================================

3 2

[libvirt-users] kvm virtual server boot order
by Yee Ja 21 Apr '11

21 Apr '11

The boot order issue http://osdir.com/ml/libvir-list/2009-08/msg00398.html resolved with the code changes specified by http://www.redhat.com/archives/libvir-list/2009-August/msg00340.html seems to suggest that hard disks, cdroms, floppy devices (since they are all disks) are ordered primarily by virtual bus type as encountered in the domain xml. This indicates that there can exist a conflict between the primary and secondary boot device types. Given the scenario where the disks of a virtual server are defined in the following order harddisk hda harddisk vda cdrom vdb cdrom hdb Boot device type is crdom. Which cdrom device will be boot attempted? >From the above links, I'm guessing that hdb will be boot attempted. Is this the actual behavior? Thanks, Yee -- yja(a)us.ibm.com * 512-286-6793 T/L 363 * Systems Management Software Development * Austin, TX

2 1

[libvirt-users] Undefined interfaces?
by Alex 19 Apr '11

19 Apr '11

Hi, I have a working fc14 install with a few virtual interfaces set up, and just noticed the following in messages: Apr 18 13:07:43 fc14 libvirtd: 13:07:43.155: error : interfaceGetXMLDesc:355 : internal error could not get interface XML description (netcf: NETLINK socket operation failed - couldn't find ifindex for interface `br0:0`) Apr 18 13:07:43 fc14 libvirtd: 13:07:43.165: error : interfaceGetXMLDesc:355 : internal error could not get interface XML description (netcf: NETLINK socket operation failed - couldn't find ifindex for interface `eth0:0`) Apr 18 13:07:43 fc14 libvirtd: 13:07:43.175: error : interfaceGetXMLDesc:355 : internal error could not get interface XML description (netcf: NETLINK socket operation failed - couldn't find ifindex for interface `eth1:0`) Apr 18 13:07:43 fc14 libvirtd: 13:07:43.185: error : interfaceGetXMLDesc:355 : internal error could not get interface XML description (netcf: NETLINK socket operation failed - couldn't find ifindex for interface `eth1:1`) I see several references that indicate this may be a bug. Or have I configured something wrong? I have some interfaces configured to be a bridge and others are configured as a regular network, with the guests being masqueraded. It's certainly possible I've configured something incorrectly. I'm not sure what information to provide to help troubleshoot this. Any information would be greatly appreciated. Thanks, Alex

2 1

[libvirt-users] Accessing a MAC address from inside the VM?
by David Lane 19 Apr '11

19 Apr '11

Greetings, I have a very simple problem, but the resolution is proving rather complex. Bare with me, I will attempt to explain. I have a host machine with no GUI. I attach an existing .img file to the host machine and need to change the MAC address (since the image was created on another machine). Is there a way from with in the guest to do this or do I have muck with the XML? And if I muck with the XML and change it, how do I pass this information up to the guest, or will the guest see it automagically or do I have to write it down and then manually transfer it into the guest's ifcfg? Using the GUI, when you change the NIC, the guest generally does not see the new MAC and grouses until you change it. Thanks. DAVID

2 2

[libvirt-users] USB2 passthrough
by Marco van Beek 18 Apr '11

18 Apr '11

Hi, I am trying to get a USB device working on a Windows 7 virtual machine, and it needs USB2 to work, so the USB passthrough does not work. I have searched for an answer and either the answer is so obvious I am being thick, or too complicated for anyone to address it. Can I simply do a PCI passthrough of a USB host controller? Would that then give me a whole load of USB2 ports I could connect to my computer? I would have to find out which port was connected to which controller, but surely it can't be that simple? Thanks, Marco -- Marco van Beek =================================================== Forget About IT Ltd. Grove Park Studios, 188-192 Sutton Court Rd, London, W4 3HR =================================================== T: 0870 757 2924 F: 0870 757 2926 M: 0788 770 3604 E: mvanbeek(a)forgetaboutit.net W: http://www.forgetaboutit.net/ =================================================== Forget About IT and FABIT are registered trademarks =================================================== Company Registered in England No. 05375739 ===================================================

1 0

[libvirt-users] Default network / net-edit
by ssaleh＠nrln.us 18 Apr '11

18 Apr '11

How do you modify the default network parameters? The net-edit command within virsh does not seem to work as expected. TIA

2 1

[libvirt-users] add file.img in libvirt
by Marco Mirabile 18 Apr '11

18 Apr '11

Good day I downloaded an image of a VMs "centOS.img" how can I open this image from libvirt? I ask you with virt-install, but I could not help me dominioYP

2 1

[libvirt-users] Problems setting up bridge
by Alex 15 Apr '11

15 Apr '11

Hi, I have a fc14 box with two interfaces that I'm having some difficulty setting up a bridge properly. I have a few kvm guests set up and running, but I've switched from using NetworkManager to using standard networking with a bridge. I've modified the ifcfg-eth0 and created ifcfg-br0 as per the standard documentation, but I think I'm having routing problems. This is what the network looks like: [root@fc14 ~]# route Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 68.195.XXX.40 * 255.255.255.248 U 0 0 0 br0 192.168.1.0 * 255.255.255.0 U 0 0 0 eth1 192.168.125.0 * 255.255.255.0 U 0 0 0 virbr0 link-local * 255.255.0.0 U 1003 0 0 eth1 link-local * 255.255.0.0 U 1004 0 0 br0 default ool-44c3c129.st 0.0.0.0 UG 0 0 0 br0 br0 Link encap:Ethernet HWaddr 40:61:86:4E:84:09 inet addr:68.195.XXX.42 Bcast:68.195.XXX.47 Mask:255.255.255.248 inet6 addr: fe80::a0e7:ecff:fefd:340a/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:3883 errors:0 dropped:0 overruns:0 frame:0 TX packets:3579 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:1818327 (1.7 MiB) TX bytes:590720 (576.8 KiB) eth0 Link encap:Ethernet HWaddr 40:61:86:4E:84:09 inet6 addr: fe80::4261:86ff:fe4e:8409/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:3891 errors:0 dropped:0 overruns:0 frame:0 TX packets:3551 errors:0 dropped:0 overruns:0 carrier:1 collisions:0 txqueuelen:1000 RX bytes:1873169 (1.7 MiB) TX bytes:588302 (574.5 KiB) Interrupt:42 eth1 Link encap:Ethernet HWaddr 00:0E:0C:3B:36:F4 inet addr:192.168.1.1 Bcast:192.168.1.255 Mask:255.255.255.0 inet6 addr: fe80::20e:cff:fe3b:36f4/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:7622 errors:0 dropped:0 overruns:0 frame:0 TX packets:6936 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:1062687 (1.0 MiB) TX bytes:4304984 (4.1 MiB) virbr0 Link encap:Ethernet HWaddr E6:ED:D2:6E:58:5C inet addr:192.168.125.1 Bcast:192.168.125.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:0 (0.0 b) TX bytes:0 (0.0 b) vnet0 Link encap:Ethernet HWaddr FE:54:00:67:2C:4C inet6 addr: fe80::fc54:ff:fe67:2c4c/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:5 errors:0 dropped:0 overruns:0 frame:0 TX packets:391 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:500 RX bytes:300 (300.0 b) TX bytes:23358 (22.8 KiB) I've left out the loopback for simplicity. The problem is that I can't reach the br0 device from a server on the eth1 device (the 192.168. network). Is this a routing issue? Should I also add the eth1 device to the bridge? If so, how would I do that? Also, can you explain under what circumstances a tun device would be used? Does that have any relation to bridging? Thanks, Alex

2 1

[libvirt-users] Where actually is the callback triggered?
by kadir yüceer 15 Apr '11

15 Apr '11

Hi all, Quick question: When the hypervisor triggers a callback which is registered for generic events, (assuming I'm connected through qemu+tcp) where exactly is the callback triggered? On the hypervisor host or on my host? Regards Kadir

2 3

[libvirt-users] custom iptables rules
by Kurian Thayil 15 Apr '11

15 Apr '11

Hi, Is there a way that we can add custom IP-Tables rules in a nat'd physical host? I need some custom rules mentioned in physical host to access some services in the guest systems. Any hints on this? Regards, Kurian Thayil.

2 1

[libvirt-users] virsh shutdown
by Zvi Dubitzky 13 Apr '11

13 Apr '11

Hi For some reason recent libvirt code (0.8.3 and even before) the 'virsh shutdown <domain> ' is not effective . It issues an ok message but the domians remains in a runnning state . Only th e destroy works fine. Any idea ? thanks Zvi Dubitzky Email:dubi@il.ibm.com

4 3

[libvirt-users] Listing shut off domains
by None 11 Apr '11

11 Apr '11

Hi, I am in the process of migrating a system that was using KVM to VMWare Workstation. The original code needs to list all the domains available on the local machine, no matter what their state is. The problem is that the shut off domains are not shown by the virsh commands on the VMWare Workstation system. On the KVM system, with one shut off domain, I get this output: maxpower@sw ~$ virsh list --all Id Name State ---------------------------------- - 00042-my-domain shut off On the VMWare Workstation system, with one shut off domain, I get this output: maxpower@sw2 ~ $ virsh -c vmwarews:///session list --all Id Name State ---------------------------------- As soon as I power on the machine in VMWare, it appears in the output as running: maxpower@sw2 ~ $ virsh -c vmwarews:///session list --all Id Name State ---------------------------------- 14063 my-domain-1 running Is this a limitation of the Workstation driver? Is there a workaround of some sort that I haven't thought of? How can I get the entire domain list when using Workstation? Thanks for your input, Alexis

1 1

[libvirt-users] Python API - setting client certificate path
by Wojciech Gumularz 09 Apr '11

09 Apr '11

Hi there, I'm trying to establish a TLS connection to libvirtd with certificates verification from a python script. My question is quite simple: is it possible to set the client certificate/keyfile path from within a python script? or it has to be in /etc/pki/libvirt/... Thanks in advance for any help Wojciech Gumularz

2 2

[libvirt-users] Can I update libvirt while KVM guest os are running
by Masanari Iida 06 Apr '11

06 Apr '11

Hello I have a simple question. Can I update libvirt RPM while KVM guest OS on RHEL5 is running. Background When I run rhn_register, then rhn-virtualization-host RPM was installed by RHN. And then libvirt-python was also installed from the RHN. The libvirt-python RPM has a dependency with libvirt. And the dependency was not only the RPM file name, and also specific RPM versions as well. # rpm -qR libvirt-python /usr/bin/python (snip) libssl.so.6()(64bit) libvirt = 0.8.2-15.el5　<== libvirt.so.0()(64bit) So the libvirt from RHEL5.5 was updated to libvirt from RHEL5.6 by this accident. If nothing was happen, I don't care. But after the force libvirt update, I couldn't do attach-disk. 　error : Failed to attach disk error : internal error missing disk device alias name for hdc Now I would like to confirm from the authority that libvirt update while running guest os is supported or not. If supported, then these messages indicated something going wrong. If not, I can ignore it. Regards, Masanari

1 2

[libvirt-users] Registering a callback in Java
by kadir yüceer 06 Apr '11

06 Apr '11

I'm able to edit org.libvirt.Domain.java and print a message when, for example suspend() is called. However I want to use eventCallback API of libvirt. Isn't this the proper piece of code to register a callback to any domain(domain parameter is null), for a lifecycle event(event id 0). conn.domainEventRegisterAny(null, 0, new VirConnectDomainEventGenericCallback() { public void eventCallback(ConnectionPointer cp, DomainPointer dp, Pointer pntr) { System.out.println("event!"); } }); I can't get the println when I suspend or resume the domains. And the reason I wrote "0" is that I could not find VIR_DOMAIN_EVENT_ID_LIFECYCLE constant in java API. Any help about where to find macros,types and enumerations in java API is appreciated. Unfortunately I can not find anything about events in javadoc neither. Any comments? Kind regards Kadir

1 0

[libvirt-users] Can not convert LVM over two vmdk files
by Marc Boorshtein 05 Apr '11

05 Apr '11

Hello, I'm trying to migrate from VMWare to KVM. I've got a redhat 5 (actually oracle unbreakable 5) vmware image that has two scsi vmdk files with logical volumes spanning both drives. Wether I convert them to raw files or load them as vmdk files, when I start them linux boots but fails while trying to initate the LVM. They are both mounted as scsi drives and in the correct order. The host is Fedora 15 Alpha. Any help would be greatly appreciated, I'd love to ditch VMWare! Thanks Marc

1 0

[libvirt-users] Protocols used for migration
by linux user 05 Apr '11

05 Apr '11

Hi all, I'd like to know what are the different protocols used for migration of KVM guests on RHEL hosts.I'm actually interested in 'virsh dump' and virsh dump executes the qemu 'migrate' command.I understand 'exec' is one of the protocols. Appreciate any pointers. For eg:(qemu) migrate -d "exec:cat >'/tmp/vm'" Thanks, Marc

2 1

[libvirt-users] Creating a network bridge with multiple interfaces
by Alex 05 Apr '11

05 Apr '11

Hi, I have an existing fedora14 system with two interfaces and several virtual interfaces (as in eth0:0, etc). They are currently configured to not be controlled by NetworkManager, but just standard network services. I've recently started working with libvirtd, so I also have a virbr0 interface as well that has its own network and is masquerading the networking for the single KVM instance I have running. I would now like to create another instance that is not masqueraded, but instead uses another virtual interface, such as eth0:1, which is an Internet-addressable IP address. When I try to do this, virt-manager seems to still configure it using NAT. I think the right way to do this is to convert the eth* instances to instead use a real bridge, correct? How can I do this? Below is the output from ifconfig on my system: eth0 Link encap:Ethernet HWaddr 40:61:86:4E:84:09 inet addr:XX.195.193.YY Bcast:XX.195.193.YY Mask:255.255.255.248 inet6 addr: fe80::4261:86ff:fe4e:8409/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:204613580 errors:0 dropped:0 overruns:0 frame:0 TX packets:220748087 errors:0 dropped:0 overruns:0 carrier:1 collisions:0 txqueuelen:1000 RX bytes:146443250035 (136.3 GiB) TX bytes:207505258199 (193.2 GiB) Interrupt:42 eth0:0 Link encap:Ethernet HWaddr 40:61:86:4E:84:09 inet addr:XX.195.193.ZZ Bcast:XX.195.193.YY Mask:255.255.255.248 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 Interrupt:42 eth1 Link encap:Ethernet HWaddr 00:0E:0C:3B:36:F4 inet addr:192.168.1.1 Bcast:192.168.1.255 Mask:255.255.255.0 inet6 addr: fe80::20e:cff:fe3b:36f4/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:241615579 errors:6 dropped:0 overruns:0 frame:3 TX packets:230218850 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:210464139370 (196.0 GiB) TX bytes:171698241573 (159.9 GiB) eth1:0 Link encap:Ethernet HWaddr 00:0E:0C:3B:36:F4 inet addr:192.168.1.2 Bcast:192.168.1.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 eth1:1 Link encap:Ethernet HWaddr 00:0E:0C:3B:36:F4 inet addr:192.168.1.100 Bcast:192.168.1.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:3439643 errors:0 dropped:0 overruns:0 frame:0 TX packets:3439643 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:849466836 (810.1 MiB) TX bytes:849466836 (810.1 MiB) virbr0 Link encap:Ethernet HWaddr 00:00:00:00:00:00 inet addr:192.168.125.1 Bcast:192.168.125.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:16 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:3088 (3.0 KiB) TX bytes:0 (0.0 b) Thanks for any ideas. Alex

1 0

[libvirt-users] iptables errors with guest
by Alex 04 Apr '11

04 Apr '11

Hi, I've just set up a fedora14 guest on a fedora14 server, and just noticed the following in messages: Apr 4 16:29:50 fc14 libvirtd: 16:29:50.767: error : virRunWithHook:857 : internal error '/sbin/iptables --table mangle --delete POSTROUTING --out-interface virbr0 --protocol udp --destination-port 68 --jump CHECKSUM --checksum-fill' exited with non-zero status 2 and signal 0: iptables v1.4.9: unknown option `--checksum-fill'#012Try `iptables -h' or 'iptables --help' for more information.#012 Apr 4 16:29:50 fc14 libvirtd: 16:29:50.840: error : virRunWithHook:857 : internal error '/sbin/iptables --table filter --delete INPUT --in-interface virbr0 --protocol udp --destination-port 69 --jump ACCEPT' exited with non-zero status 1 and signal 0: iptables: Bad rule (does a matching rule exist in that chain?).#012 Apr 4 16:29:50 fc14 libvirtd: 16:29:50.990: error : virRunWithHook:857 : internal error '/sbin/iptables --table mangle --insert POSTROUTING --out-interface virbr0 --protocol udp --destination-port 68 --jump CHECKSUM --checksum-fill' exited with non-zero status 2 and signal 0: iptables v1.4.9: unknown option `--checksum-fill'#012Try `iptables -h' or 'iptables --help' for more information.#012 Apr 4 16:29:50 fc14 libvirtd: 16:29:50.990: warning : networkAddIptablesRules:850 : Could not add rule to fixup DHCP response checksums on network 'default'. Apr 4 16:29:50 fc14 libvirtd: 16:29:50.990: warning : networkAddIptablesRules:851 : May need to update iptables package & kernel to support CHECKSUM rule. Is this something I'm doing wrong or is it a known problem? I have a basic understanding of the iptables rules that libvirtd, but not enough to troubleshoot this. What other information can I provide to help troubleshoot this? Thanks, Alex

1 0

[libvirt-users] Release of libvirt-0.9.0
by Daniel Veillard 04 Apr '11

04 Apr '11

As scheduled, libvirt 0.9.0 was tagged and pushed today, it's available from FTP at: ftp://libvirt.org/libvirt/ This is a large release w.r.t. the amount of features and changes, and well worth bumping the middle version number. We are also getting closer to a 1.0.0 release ! Features: - Support cputune cpu usage tuning (Osier Yang and Nikunj A. Dadhania) - Add public APIs for storage volume upload/download (Daniel P. Berrange) - Add public API for setting migration speed on the fly (Daniel P. Berrange) - Add libxenlight driver (Jim Fehlig and Markus Groß) - qemu: support migration to fd (Eric Blake) - libvirt: add virDomain{Get,Set}BlkioParameters (Gui Jianfeng) - setmem: introduce a new libvirt API (virDomainSetMemoryFlags) (Taku Izumi) - Expose event loop implementation as a public API (Daniel P. Berrange) - Dump the debug buffer to libvirtd.log on fatal signal (Daniel Veillard) - Audit support (Eric Blake) Documentation: - fix typo (Eric Blake) - correct invalid xml (Eric Blake) - virsh: Fix documentation for memtune command (Jiri Denemark) - Fix several formatting mistakes in doc (Michal Privoznik) - mention C89 syntax preferences (Eric Blake) - document recent hook additions (Eric Blake) - Update on the goal page (Daniel Veillard) - Document first release with spice and qxl (Cole Robinson) - Add schema definition for imagelabel (Osier Yang) - update virGetVersion description (Tiziano Mueller) - Improve logging documentation including the debug buffer (Daniel Veillard) - update windows page for initial libvirt 0.8.8 installer (Justin Clift) - formatdomain.html.in: Fix spelling PIC->PCI (Philipp Hahn) - fix missing </p> (Eric Blake) - documenting the 802.1Qbg parameters of a 'direct' interface (Gerhard Stenzel) - silence warnings about generated API docs (Eric Blake) - document <driver name='vhost'/> for interfaces (Eric Blake) - correct range of default NAT subnet (Eric Blake) - formatdomain: Add release info for disk <driver> attributes (Cole Robinson) - Fix spelling mistake: seek (Philipp Hahn) - maint: fix grammar in error message (Eric Blake) Portability: - virsh: fix mingw failure on creating nonblocking pipe (Eric Blake) - Remove iohelper on Win32 since it is not required (Daniel P. Berrange) - Fix domain events C example on Win32 (Daniel P. Berrange) - build: fix compilation on mingw (Eric Blake) - util: use SCM_RIGHTS in virFileOperation when needed (Eric Blake) - Don't use INT64_MAX in libvirt.h because it requires stdint.h (Matthias Bolte) - libvirtd: Remove indirect linking (Guido Günther) - build: avoid compiler warning on cygwin (Eric Blake) - build: fix build on cygwin (Eric Blake) - build: fix building error when building without libvirtd (Wen Congyang) - virsh: Remove indirect link against libxml2 (Guido Günther) - Fix build on cygwin (Daniel Veillard) - Add check for kill() to fix build of cgroups on win32 (Daniel P. Berrange) - build: fix broken mingw cross-compilation (Eric Blake) Bug fixes: - fix memory leak in qemuProcessHandleGraphics() (Wen Congyang) - do not lock vm while allocating memory (Wen Congyang) - Fix libxl driver startup (Daniel Veillard) - qemu: Ignore libvirt debug messages in qemu log (Jiri Denemark) - qemu: Fix improper logic of qemuCgroupSetup (Osier Yang) - free tmp after unlinking it (Wen Congyang) - qemu: Fix media eject with qemu-0.12.* (Jiri Denemark) - check whether qemuMonitorJSONHMP() failed (Wen Congyang) - do not send monitor command after monitor meet error (Wen Congyang) - qemu: unlock qemu driver before return from domain save (Hu Tao) - qemu: fix regression with fd labeling on migration (Eric Blake) - Ignore return value of virDomainObjUnref (Markus Groß) - Fix infinite loop in daemon if client quits with multiple streams open (Daniel P. Berrange) - qemu: fix regression that hangs on save failure (Eric Blake) - qemu: fix restoring a compressed save image (Eric Blake) - util: allow clearing cloexec bit (Eric Blake) - logging: always NUL-terminate circular buffer (Eric Blake) - tests: don't alter state in $HOME (Eric Blake) - qemu: don't restore state label twice (Eric Blake) - Fix syntax error in configure.ac (Osier Yang) - remote: Don't leak gnutls session on negotiation error (Matthias Bolte) - hooks: fix regression in previous patch (Eric Blake) - Add missing { for qemudDomainInterfaceStats (Philipp Hahn) - daemon: Avoid resetting errors before they are reported (Jiri Denemark) - fix the check of the output of monitor command 'device_add' (Wen Congyang) - Make error reporting in libvirtd thread safe (Jiri Denemark) - update domain status forcibly even if attach a device failed (Wen Congyang) - util: Fix return value for virJSONValueFromString if it fails (Osier Yang) - Initialization error of qemuCgroupData in Qemu host usb hotplug (Wen Congyang) - build: fix missing initializer (Eric Blake) - Fix uninitialized variable & error reporting in LXC veth setup (Daniel P. Berrange) - udev: fix regression with qemu:///session (Eric Blake) - logging: fix off-by-one bug (Eric Blake) - do not report OOM error when prepareCall() failed (Wen Congyang) - Don't return an error on failure to create blkio controller (Hu Tao) - qemu: respect locking rules (Eric Blake) - openvz: fix a simple bug in openvzListDefinedDomains() (Jean-Baptiste Rouault) - Fix delayed event delivery when SASL is active (Daniel P. Berrange) - qemu: Fix copy&paste error messages in text monitor (Jiri Denemark) - do not unref obj in qemuDomainObjExitMonitor* (Wen Congyang) - qemu: check driver name while attaching disk (Wen Congyang) - remote: Add missing virCondDestroy calls (Matthias Bolte) - build: improve rpm generation for distro backports (Eric Blake) - storage: Fix a problem which will cause libvirtd crashed (Osier Yang) - Fix misc bugs in virCommandPtr (Daniel P. Berrange) - libvirt: fix a simple bug in virDomainSetMemoryFlags() (Taku Izumi) - qemu: Check the unsigned integer overflow (Osier Yang) - audit: eliminate potential null pointer deref when auditing macvtap devices (Laine Stump) - network driver: don't send default route to clients on isolated networks (Laine Stump) - virsh: Free stream when shutdown console (Osier Yang) - Add missing checks for read only connections (Guido Günther) - qemu: fix -global argument usage (Eric Blake) - Make sure we reset the umask on the error path (Guido Günther) - qemu: Stop guest CPUs before creating a snapshot (Jiri Denemark) - qemu: Escape snapshot name passed to {save,load,del}vm (Jiri Denemark) - qemu: Fix warnings in event handlers (Jiri Denemark) - storage: Update qemu-img flag checking (Osier Yang) - Make sure the rundir is accessible by the user (Guido Günther) - Fix a wrong error message thrown to user (Hu Tao) - unlock eventLoop before calling callback function (Wen Congyang) - fixes for several memory leaks (Phil Petty) - Fix a counter bug in the log buffer (Daniel Veillard) - qemu: avoid corruption of domain hashtable and misuse of freed domains (Laine Stump) - qemu: Add missing lock of virDomainObj before calling virDomainUnref (Laine Stump) - esx: Escape password for XML (Matthias Bolte) - util: correct retry path in virFileOperation (Eric Blake) - util: Allow removing hash entries in virHashForEach (Jiri Denemark) - qemu: avoid double close on domain restore (Eric Blake) - Fix port value parsing for serial and parallel ports (Michal Novotny) - Fix off-by-1 in virFileAbsPath. (Daniel P. Berrange) - security: avoid memory leak (Eric Blake) - protect the scsi controller to be deleted when it is in use (Wen Congyang) - virsh: freecell --all getting wrong NUMA nodes count (Michal Privoznik) - remove duplicated call to reportOOMError (Christophe Fergeau) Improvements: - Make check_fc_host() and check_vport_capable() usable as rvalues (Guido Günther) - maint: avoid locale-sensitivity in string case comparisons (Eric Blake) - extend logging to record configuration-related changes (Naoya Horiguchi) - Add libvirt_iohelper to spec file (Daniel Veillard) - cputune: New tests for cputune XML (Osier Yang) - cputune: Support cputune for xend driver (Osier Yang) - cputune: Support cputune for lxc driver (Osier Yang) - cputune: Support cputune for qemu driver (Osier Yang) - cputune: Implementations of parsing and formating cputune xml (Osier Yang) - cputune: Add data structures presenting cputune XML (Osier Yang) - cputune: Add document for cputune XML (Osier Yang) - cputune: Add XML schema for cputune xml (Osier Yang) - qemu: improve error message on failed fd transfer (Eric Blake) - maint: ignore new built file (Eric Blake) - Add domainSuspend/Resume to libxl driver (Markus Groß) - Add domainGetOSType to libxl driver (Markus Groß) - Add domainGetSchedulerType to libxl driver (Markus Groß) - Implements domainXMLTo/FromNative in libxl driver (Markus Groß) - Add vcpu functions to libxl driver (Markus Groß) - List authors in copyright headers (Markus Groß) - Add event callbacks to libxl driver (Markus Groß) - Remote protocol support for storage vol upload/download APIs (Daniel P. Berrange) - Support volume data upload/download APIs in storage driver (Daniel P. Berrange) - Add vol-upload and vol-download commands to virsh (Daniel P. Berrange) - Enhance the streams helper to support plain file I/O (Daniel P. Berrange) - Update event loop example programs to demonstrate best practice (Daniel P. Berrange) - qemu: support fd: migration with compression (Eric Blake) - qemu: skip granting access during fd migration (Eric Blake) - qemu: consolidate migration to file code (Eric Blake) - qemu: use common API for reading difficult files (Eric Blake) - qemu, storage: improve type safety (Eric Blake) - util: adjust indentation in previous patch (Eric Blake) - util: rename virFileOperation to virFileOpenAs (Eric Blake) - storage: simplify fd handling (Eric Blake) - qemu: simplify domain save fd handling (Eric Blake) - qemu: allow simple domain save to use fd: protocol (Eric Blake) - Update of localisations, switch to transifex (Daniel Veillard) - build: shorten libxenlight summary for consistent alignment (Eric Blake) - command: add virCommandAbort for cleanup paths (Eric Blake) - command: don't mix RunAsync and daemons (Eric Blake) - command: properly diagnose process exit via signal (Eric Blake) - Add memory functions to libxl driver (Markus Groß) - build: enforce reference count checking (Eric Blake) - maint: prohibit access(,X_OK) (Eric Blake) - Get cpu time and current memory balloon from libxl (Markus Groß) - build: nuke all .x-sc* files, and fix VPATH syntax-check (Eric Blake) - command: reject pidfile on non-daemon (Eric Blake) - rpm: add missing dependencies (Eric Blake) - rpm: separate runtime and build requirements (Eric Blake) - qemu: simplify monitor callbacks (Eric Blake) - 8021Qbh: use preassociate-rr during the migration prepare stage (Roopa Prabhu) - Wire up virDomainMigrateSetSpeed into QEMU driver (Daniel P. Berrange) - Wire up virDomainMigrateSetSpeed for the remote RPC driver (Daniel P. Berrange) - maint: update authors (Eric Blake) - Disable libxl build in RPM on Fedora < 16 (Daniel P. Berrange) - qemu: fallback to HMP drive_add/drive_del (Hu Tao) - qemu: Only use HMP passthrough if it is supported (Jiri Denemark) - qemu: Detect support for HMP passthrough (Jiri Denemark) - qemu: add two hook script events "prepare" and "release" (Thibault Vincent) - qemu: simplify interface fd handling in monitor (Eric Blake) - qemu: simplify PCI configfd handling in monitor (Eric Blake) - qemu: simplify monitor fd error handling (Eric Blake) - util: guarantee sane errno in virFileIsExecutable (Eric Blake) - Don't build libxenlight driver for Xen 4.0 (Jim Fehlig) - network driver: log error and abort network startup when radvd isn't found (Laine Stump) - build: translate changes in previous patch (Eric Blake) - Ensure binary is resolved wrt $PATH in virExec (Daniel P. Berrange) - util: Forbid calling hash APIs from iterator callback (Jiri Denemark) - Avoid taking lock in libvirt debug dump (Daniel Veillard) - unlock the monitor when unwatching the monitor (Wen Congyang) - Add vim configuration that makes vim auto-indent code (Hu Tao) - virsh: fix memtune's help message for swap_hard_limit (Nikunj A. Dadhania) - Add PCI sysfs reset access (Alex Williamson) - Support Xen sysctl v8, domctl v7 (Jim Fehlig) - macvtap: log an error if on failure to connect to netlink socket (Laine Stump) - qemu: improve efficiency of dd during snapshots (Eric Blake) - virsh: allow empty string arguments (Eric Blake) - qemu: Fallback to HMP when cpu_set QMP command is not found (Wen Congyang) - Change message for VIR_FROM_RPC error domain (Daniel P. Berrange) - Add compat function for geteuid() (Daniel P. Berrange) - Add virSetBlocking() to allow O_NONBLOCK to be toggle on or off (Daniel P. Berrange) - qemu: use more appropriate error (Eric Blake) - Make LXC container startup/shutdown/I/O more robust (Daniel P. Berrange) - Allow to dynamically set the size of the debug buffer (Daniel Veillard) - qemu: consolidate duplicated monitor migration code (Eric Blake) - qemu: use lighter-weight fd:n on incoming tunneled migration (Eric Blake) - Fix performance problem of virStorageVolCreateXMLFrom() (Minoru Usui) - libvirt-guests: avoid globbing when splitting $URIS (Eric Blake) - libvirt-guest.init: quoting variables (Philipp Hahn) - virsh: Insert error messages to avoid a quiet abortion of commands (Michal Privoznik) - python: Use hardcoded python path in libvirt.py (Jiri Denemark) - virsh: Allow starting domains by UUID (Jiri Denemark) - network driver: Use a separate dhcp leases file for each network (Laine Stump) - network driver: Start dnsmasq even if no dhcp ranges/hosts are specified. (Laine Stump) - libvirt-guest.init: handle domain name with spaces (Philipp Hahn) - domain.rng vs. formatdomain.html#elementsUSB (Philipp Hahn) - Ignore backing file errors in FS storage pool (Philipp Hahn) - remote-protocol: implement new BlkioParameters API (Gui Jianfeng) - virsh: Adding blkiotune command to virsh tool (Gui Jianfeng) - qemu: implement new BlkioParameters API (Gui Jianfeng) - libvirt: implements virDomain{Get,Set}BlkioParameters (Gui Jianfeng) - setmem: add the new options to "virsh setmem" command (Taku Izumi) - setmem: implement the remote protocol to address the new API (Taku Izumi) - setmem: implement the code to address the new API in the qemu driver (Taku Izumi) - audit: audit use of /dev/net/tun, /dev/tapN, /dev/vhost-net (Eric Blake) - qemu: don't request cgroup ACL access for /dev/net/tun (Eric Blake) - qemu: support vhost in attach-interface (Eric Blake) - qemu: Refactor qemuDomainSnapshotCreateXML (Jiri Denemark) - qemu: Fallback to HMP for snapshot commands (Jiri Denemark) - qemu: Setup infrastructure for HMP passthrough (Jiri Denemark) - qemu: Replace deprecated option of qemu-img (Osier Yang) - audit: also audit cgroup ACL permissions (Eric Blake) - cgroup: allow fine-tuning of device ACL permissions (Eric Blake) - audit: rename remaining qemu audit functions (Eric Blake) - audit: also audit cgroup controller path (Eric Blake) - audit: split cgroup audit types to allow more information (Eric Blake) - audit: tweak audit messages to match conventions (Eric Blake) - Don't overwrite virRun error messages (Cole Robinson) - virsh: Change option parsing functions to return tri-state information (Michal Privoznik) - virsh: change vshCommandOptString return type and fix const-correctness (Michal Privoznik) - support to detach USB disk (Wen Congyang) - rename qemuDomainDetachSCSIDiskDevice to qemuDomainDetachDiskDevice (Wen Congyang) - qemu_hotplug: Reword error if spice password change not available (Cole Robinson) - Move event code out of the daemon/ into src/util/ (Daniel P. Berrange) - Convert daemon/virsh over to use primary event APIs, rather than impl (Daniel P. Berrange) - Cleaning up some of the logging code (Daniel Veillard) - qemu: Support vram for video of qxl type (Osier Yang) - Add an an internal API for emergency dump of debug buffer (Daniel Veillard) - Add logrotate support for libvirtd.log (Daniel Veillard) - Change default log policy to libvirtd.log instead of syslog (Daniel Veillard) - Force all logs to go to the round robbin memory buffer (Daniel Veillard) - AUTHORS: adjust to preferred spelling (KAMEZAWA Hiroyuki) - Pass virSecurityManagerPtr to virSecurityDAC{Set, Restore}ChardevCallback (Soren Hansen) - maint: update to latest gnulib (Eric Blake) - Attempt to improve an error message (Daniel P. Berrange) - add additional event debug points (Daniel P. Berrange) - qemu: only request sound cgroup ACL when required (Eric Blake) - Add support for multiple serial ports into the Xen driver (Michal Novotny) - Add APIs for killing off processes inside a cgroup (Daniel P. Berrange) - Allow hash tables to use generic pointers as keys (Daniel P. Berrange) - Remove deallocator parameter from hash functions (Daniel P. Berrange) - Make commandtest more robust wrt its execution environment (Daniel P. Berrange) - audit: audit qemu pci and usb device passthrough (Eric Blake) - audit: audit qemu memory and vcpu adjusments (Eric Blake) - audit: add qemu hooks for auditing cgroup events (Eric Blake) - audit: prepare qemu for listing vm in cgroup audits (Eric Blake) - cgroup: determine when skipping non-devices (Eric Blake) - virExec: avoid uninitialized memory usage (Eric Blake) - Allow 32-on-64 execution for LXC guests (Daniel P. Berrange) - Put <stdbool.h> into internal.h so it is available everywhere (Daniel P. Berrange) - qemu: Switch over command line capabilities to virBitmap (Jiri Denemark) - qemu: Rename qemud\?CmdFlags to qemuCaps (Jiri Denemark) - qemu: Use helper functions for handling cmd line capabilities (Jiri Denemark) - qemu: Rename QEMUD_CMD_FLAG_* to QEMU_CAPS_* (Jiri Denemark) - util: Add API for converting virBitmap into printable string (Jiri Denemark) - util: Use unsigned long as a base type for virBitmap (Jiri Denemark) - Expose name + UUID to LXC containers via env variables (Daniel P. Berrange) - Fix discard of expected errors (Daniel P. Berrange) - Fix group/mode for /dev/pts inside LXC container (Daniel P. Berrange) - 802.1Qbh: Delay IFF_UP'ing interface until migration final stage (Roopa Prabhu) - storage: make debug log more useful (Osier Yang) - virsh: replace vshPrint with vshPrintExtra for snapshot list Otherwise extra information will be printed even if "--quiet" is specified. (Osier Yang) - check device-mapper when building with mpath or disk storage driver (Wen Congyang) - build: add dependency on gnutls-utils (Eric Blake) - Renamed functions in xenxs (Markus Groß) - Moved XM formatting functions to xenxs (Markus Groß) - Moved XM parsing functions to xenxs (Markus Groß) - Moved SEXPR formatting functions to xenxs (Markus Groß) - Moved SEXPR parsing functions to xenxs (Markus Groß) - Moved some SEXPR functions from xen-unified (Markus Groß) - Moved SEXPR unit to utils (Markus Groß) - virt-*-validate.in: quote all variable references (Dan Kenigsberg) - virt-pki-validate: behave when CERTTOOL is missing (Dan Kenigsberg) - autobuild.sh: use VPATH build (Eric Blake) - maint: fix 'make dist' in VPATH build (Eric Blake) - build: don't require pod2man for tarball builds (Eric Blake) - hash: make virHashFree more free-like (Eric Blake) - build: Fix API docs generation in VPATH build (Jiri Denemark) - Remove all object hashtable caches from virConnectPtr (Daniel P. Berrange) - nwfilter: enable rejection of packets (Stefan Berger) - Drop empty argument from dnsmasq call (Guido Günther) - esx: Ignore malformed host UUID from BIOS (Matthias Bolte) - build: speed up non-maintainer builds (Eric Blake) - build: recompute symbols after changing configure options (Eric Blake) - Requires gettext for client package (Osier Yang) - Do not add drive 'boot=on' param when a kernel is specified (Jim Fehlig) - factor common code in virHashAddEntry and virHashUpdateEntry (Christophe Fergeau) - add hash table rebalancing in virHashUpdateEntry (Christophe Fergeau) - hash: modernize debug code (Eric Blake) - build: improve 'make install' for VPATH builds (Eric Blake) - check more error info about whether drive_add failed (Wen Congyang) - logging: make VIR_ERROR and friends preserve errno (Eric Blake) - maint: avoid 'make syntax-check' from tarball (Eric Blake) - Give each virtual network bridge its own fixed MAC address (Laine Stump) - Allow brAddTap to create a tap device that is down (Laine Stump) - Add txmode attribute to interface XML for virtio backend (Laine Stump) - Restructure domain struct interface "driver" data for easier expansion (Laine Stump) - build: Fix VPATH build (Jiri Denemark) - storage: Allow to delete device mapper disk partition (Osier Yang) Cleanups: - The next release is 0.9.0 not 0.8.9 (Daniel Veillard) - maint: use space, not tab, in remote_protocol-structs (Eric Blake) - Remove the Open Nebula driver (Daniel P. Berrange) - domain_conf: drop unused ref-count in snapshot object (Eric Blake) - Update the set of maintainers for the project (Daniel Veillard) - Make virDomainObjParseNode() static (Hu Tao) - maint: make spacing in .sh files easier (Eric Blake) - network driver: Fix indentation from previous commit (Laine Stump) - qemu: Rename qemuMonitorCommandWithHandler as qemuMonitorText* (Jiri Denemark) - qemu: Rename qemuMonitorCommand{,WithFd} as qemuMonitorHMP* (Jiri Denemark) - maint: avoid long lines in more tests (Eric Blake) - maint: kill all remaining uses of old DEBUG macro (Eric Blake) - maint: Expand tabs in python code (Jiri Denemark) - remove space between function name and ( (Christophe Fergeau) - don't check for NULL before calling virHashFree (Christophe Fergeau) - remove no longer needed calls to virReportOOMError (Christophe Fergeau) - Move all the QEMU migration code to a new file (Daniel P. Berrange) - Split all QEMU process mangement code into separate file (Daniel P. Berrange) Thanks to everybody who helped one way or another for this release ! Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

[libvirt-users] LIBVIRT migration
by Marcela Castro León 04 Apr '11

04 Apr '11

Hello I''m developing a system using libvirt and I need to use "migrate" to move guest to different hosts. I'm testing using a cluster connected by ssh, and I'm connect ok trought them, using virsh and virt-manager. But if I try to migrate doing: *virsh # migrate --live scompi1 qemu+ssh://rionegro/system* I got: *error: monitor socket did not show up.: Connection refused* I've try using the virt-manager, but troughs: *Unable to migrate guest: Traceback (most recent call last): File "/usr/share/virt-manager/**virtManager/migrate.py", line 453, in _async_migrate vm.migrate(dstconn, migrate_uri, rate, live, secure) File "/usr/share/virt-manager/**virtManager/domain.py", line 230, in migrate self._backend.migrate(**destconn.vmm, flags, newname, interface, rate) File "/usr/lib/python2.6/dist-**packages/libvirt.py", line 429, in migrate if ret is None:raise libvirtError('**virDomainMigrate() failed', dom=self) libvirtError: monitor socket did not show up.: Connection refused* Someone can help me? Thank you very much. Marcela.

4 7

[libvirt-users] bridge and route without IP
by Robert Šmol 01 Apr '11

01 Apr '11

Hi, I want to forward all traffic received on eth1 into virtual domain 'firewall' (where I want to configure public IP+shorewall) I've created following network: node002 qemu # cat networks/int-001.xml <network> <name>int-001</name> <bridge name="intbr0" /> <forward mode="route" device="eth1"/> </network> virsh # net-create networks/int-001.xml error: Failed to create network from networks/int-001.xml error: internal error Forwarding requested, but no IP address provided But wan to set the public IP on the Firewall side. How do I do that? Would the macvtap help me? Regards, Robert -- stereoIT - Open Source Technology Partners http://www.stereoit.com mobile: +420 776 762378

2 1

[libvirt-users] Using Restore
by Marcela Castro León 01 Apr '11

01 Apr '11

Hello: I'm developing a fault tolerance system, and I'm trying to use save/restore operation. But it's working fine when I restore the file saved on the same guest, but If I move it to another host, and then try to restore using the file previously saved I got this error. I ask for help to know if this kind of operation is allows, and if it is, how can I do to work it. The exactly sequence using virsh I've test is: On [HOST1]: Using virsh 1) save [domain] [file] 2) restore file 3) destroy [domain] On [HOST1] using ubuntu sh 4) cp [guest.img] [guest.xml] [file] to HOST2 On [HOST2] using virsh 5) define [guest.xml] (using image on destination in HOST2) 6) restore [file] The restore troughs the following message: *virsh # restore sv-chubut-2011-04-01-09:58 error: Failed to restore domain from sv-chubut-2011-04-01-09:58 error: monitor socket did not show up.: Connection refused* Thank you very much. Marcela.

1 0

[libvirt-users] ANNOUNCE: oz 0.3.0 release
by Chris Lalancette 31 Mar '11

31 Mar '11

All, I'm pleased to announce release 0.3.0 of Oz. Oz is a program for doing automated installation of guest operating systems with limited input from the user. Release 0.3.0 is mostly a bugfix release for Oz. However, there are also a few minor features included as well. Some of the highlights between Oz 0.2.0 and 0.3.0 are: - Add the ability to specify the output directory for screenshots (the default is still CWD). - Add the ability to specify the bridge to use for installs (the default is still the first libvirt NAT bridge on the system). - Nicer error messages from the command-line tools when debugging is off. - Add the ability to install Debian 5 and 6. - Detect servers that can't possibly work for installing RedHat/Fedora OSs, and fail the install. - Allow installs with spaces in the name. - Add the ability to install OpenSUSE 11.4. More documentation on Oz is available here: http://aeolusproject.org/oz.html If you have any questions or comments about oz, please feel free to contact aeolus-devel(a)fedorahosted.org or me (clalance(a)redhat.com) directly. Thanks, -- Chris Lalancette

1 0