- Devel - libvirt List Archives

[libvirt] [PATCH 0/9] virDomain{Attach,Detach}DeviceFlags patches
by Jim Fehlig 09 Feb '10

09 Feb '10

This set implements two new APIs, virDomainAttachDeviceFlags and virDomainDetachDeviceFlags as discussed here https://www.redhat.com/archives/libvir-list/2009-December/msg00124.html Introduce two new APIs virDomainAttachDeviceFlags(virDomainPtr dom, const char *xml, unsigned flags) virDomainDetachDeviceFlags(virDomainPtr dom, const char *xml, unsigned flags) with flags being one or more from VIR_DOMAIN_DEVICE_MODIFY_CURRENT, VIR_DOMAIN_DEVICE_MODIFY_LIVE, VIR_DOMAIN_DEVICE_MODIFY_CONFIG. If caller specifies CURRENT (default), add or remove the device depending on the current state of domain. E.g. if domain is active add or remove the device to/from live domain, if it is inactive change the persistent config. If caller specifies LIVE, only change the active domain. If caller specifies CONFIG, only change persistent config - even if the domain is active. If caller specifies both LIVE and CONFIG, then change both. If a driver can not satisfy the exact requested flags it must return an error. E.g if user specified LIVE but the driver can only change live and persisted config, the driver must fail the request. The existing virDomain{Attach,Detach}Device is now explicitly restricted to active domains and is equivalent to virDomain{Attach,Detach}DeviceFlags(LIVE). Finally, virsh {attach,detach}-{disk,interface,device} has been modified to add a --persistent flag in order to set the appropriate flags when calling the new APIs. Jim Fehlig (9): Restrict virDomain{Attach,Detach}Device to active domains Public API Internal API Public API Implementation Wire protocol format Remote driver Server side dispatcher domain{Attach,Detach}DeviceFlags handler for drivers Modify virsh commands daemon/remote.c | 53 ++++++++++++++++++ include/libvirt/libvirt.h.in | 13 +++++ src/driver.h | 10 ++++ src/esx/esx_driver.c | 2 + src/libvirt.c | 120 ++++++++++++++++++++++++++++++++++++++---- src/libvirt_public.syms | 6 ++ src/lxc/lxc_driver.c | 2 + src/opennebula/one_driver.c | 2 + src/openvz/openvz_driver.c | 2 + src/phyp/phyp_driver.c | 2 + src/qemu/qemu_driver.c | 26 +++++++++ src/remote/remote_driver.c | 54 +++++++++++++++++++ src/remote/remote_protocol.x | 17 ++++++- src/test/test_driver.c | 2 + src/uml/uml_driver.c | 2 + src/vbox/vbox_tmpl.c | 24 ++++++++ src/xen/proxy_internal.c | 4 +- src/xen/xen_driver.c | 42 +++++++++++++-- src/xen/xen_driver.h | 4 +- src/xen/xen_hypervisor.c | 4 +- src/xen/xen_inotify.c | 4 +- src/xen/xend_internal.c | 98 ++++++++++++++++++++++++++++------ src/xen/xm_internal.c | 30 +++++++---- src/xen/xs_internal.c | 4 +- tools/virsh.c | 55 +++++++++++++++++-- 25 files changed, 523 insertions(+), 59 deletions(-)

3 33

[libvirt] [PATCH] Create raw storage files with O_DSYNC
by Jiri Denemark 09 Feb '10

09 Feb '10

When creating preallocated large raw files opening them with O_DSYNC prevents long delays in reading because cache pages can be immediately reused without writing them on a disk first. Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- src/storage/storage_backend.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/storage/storage_backend.c b/src/storage/storage_backend.c index e79294a..8aab9ab 100644 --- a/src/storage/storage_backend.c +++ b/src/storage/storage_backend.c @@ -301,7 +301,7 @@ virStorageBackendCreateRaw(virConnectPtr conn, goto cleanup; } - if ((fd = open(vol->target.path, O_RDWR | O_EXCL)) < 0) { + if ((fd = open(vol->target.path, O_RDWR | O_EXCL | O_DSYNC)) < 0) { virReportSystemError(conn, errno, _("cannot open new path '%s'"), vol->target.path); -- 1.6.6.1

2 2

[libvirt] ANNOUNCE: New releases virt-manager 0.8.3 and virtinst 0.500.2
by Cole Robinson 09 Feb '10

09 Feb '10

I'm happy to announce two new releases: virt-manager 0.8.3: virt-manager is a desktop application for managing KVM and Xen virtual machines via libvirt. virtinst 0.500.2: virtinst is a collection of command line tools for provisioning libvirt virtual machines, including virt-install and virt-clone. The releases can be downloaded from: http://virt-manager.org/download.html The direct download links are: http://virt-manager.org/download/sources/virt-manager/virt-manager-0.8.3.ta… http://virt-manager.org/download/sources/virtinst/virtinst-0.500.2.tar.gz The virt-manager release includes: - New ability to manage network interfaces: start, stop, and view existing interfaces. Provision new bridge, bond, and vlan devices. - New option to 'customize VM before install', which allows adjusting most VM options from the install wizard. The virtinst release includes: - virt-install: --autostart option for setting domain autostart flag - virt-install: Allow specifying --host-device with lsusb and lspci format (0x1234:0x5678 for USB, etc) - virtinst: New 'Interface' module for building libvirt interface XML Thanks to everyone who has contributed to this release through testing, bug reporting, submitting patches, and otherwise sending in feedback! Thanks, Cole

1 0

Re: [libvirt] Live migration woes
by Lance Albertson 09 Feb '10

09 Feb '10

> On Fri, Feb 05, 2010 at 03:21:32PM -0500, Thomas Sjolshagen wrote: >>> is anyone using libvirt-git and qemu-kvm post-0.11 and can live >>> migrate VMs successfully for- and backwards? qemu-kvm-0.11 seems >>> to be the latest release were live migration works, and I'm >>> trying to figure out if its a kvm or libvirt problem, since there >>> were other things (balloon parameter f.e.) that were changed with >>> qemu-kvm-0.12 and caused troubles when using libvirt. >> >> Yes. >> >> I'm using the virt-preview repo (rawhide) and I just noticed, >> today, after re-upgrading to 0.12.2-5 (since -4 has a rather nasty >> IO problem for virtio devices against raw image files) that my >> live-migrations fail with the following error message on the >> destination system (i.e. the system the I'm trying to migrate to): > >> /var/log/libvirt/qemu/<guest>.log: Unknown savevm section >> -2038417646 load of migration failed > > This looks like a new QEMU bug to me I can confirm this bug is happening on Ganeti as well [1]. It appears that if you sync the time on the guest VM before migrating back to the other node it will "succeed" most of the time. Has anyone opened a bug for this on the qemu-kvm bug tracker yet? I was about to do that myself. Thanks- [1] http://groups.google.com/group/ganeti/browse_thread/thread/9b2c556557e749cf -- Lance Albertson lance <at> osuosl.org Systems Administrator / Architect Open Source Lab Network Services Oregon State University Work: 541-737-9975 PGP Key: 0x27F4B742 GPG Fingerprint 0423 92F3 544A 1282 5AB1 4D07 416F A15D 27F4 B742

2 1

[libvirt] Inbound NAT and iptables rules
by Karl Vogel 09 Feb '10

09 Feb '10

This issue has been brought up a few times, but I haven't found any real solution yet. The problem is with the way libvirt adds iptables rules. I have a KVM Host with 1 public IP address, on which a few virtual machines run. I want to be able to direct certain ports on the public IP to certain virtual machine's internal IP. For example: allow conneciton to 'public-ip' tcp port: 80 to 'vm1' Which is possible using following rules: $ iptables -I FORWARD 1 -d vm1 -p tcp --dport 80 -j ACCEPT $ iptables -t nat -A PREROUTING -p tcp -d public-ip --dport 80 -j DNAT --to-destination vm1 However the problem is that libvirt _always_ adds it's rules to the top of the FORWARD rules, hence it blocks any 'custom' rules to the VM machines. ie. Chain FORWARD (policy ACCEPT 0 packets, 0 bytes) pkts bytes target prot opt in out source destination 39 2964 ACCEPT all -- * virbr0 0.0.0.0/0 192.168.122.0/24 state RELATED,ESTABLISHED 39 2964 ACCEPT all -- virbr0 * 192.168.122.0/24 0.0.0.0/0 2 656 ACCEPT all -- virbr0 virbr0 0.0.0.0/0 0.0.0.0/0 2 120 REJECT all -- * virbr0 0.0.0.0/0 0.0.0.0/0 reject-with icmp-port-unreachable 0 0 REJECT all -- virbr0 * 0.0.0.0/0 0.0.0.0/0 reject-with icmp-port-unreachable Only established connections are allowed with the libvirt rules, hence the forwarded connection to port 80 will be rejected by the 4th libvirt rule. The only way to make it work is to insert the rule _after_ libvirt has started the network.. and if the libvirt network ever gets restarted, the rule will be disabled since libvirt inserts it's rules at the top of the chain. Is there a proper way to achieve this with libvirt ?!

1 1

[libvirt] qemu can't parse floppy cmdline
by Cole Robinson 09 Feb '10

09 Feb '10

Hi Dan, With qemu-0.12.2-5.fc13.x86_64 in fedora, libvirt is generating a cmdline which qemu-kvm can't parse. XML is: <domain type='kvm'> <name>winxp_32</name> <uuid>3485e26d-c370-c0b0-32e2-a041a12f5cfc</uuid> <memory>1048576</memory> <currentMemory>1048576</currentMemory> <vcpu>1</vcpu> <os> <type arch='x86_64' machine='pc-0.11'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> <pae/> </features> <clock offset='localtime'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/bin/qemu-kvm</emulator> <disk type='file' device='disk'> <driver name='qemu'/> <source file='/mnt/devel/vms/disks/winxp_32.qcow2.img'/> <target dev='hda' bus='ide'/> <address type='drive' controller='0' bus='0' unit='0'/> </disk> <disk type='file' device='cdrom'> <driver name='qemu'/> <source file='/mnt/devel/media/win_2003_64.iso'/> <target dev='hdc' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' unit='0'/> </disk> <disk type='file' device='cdrom'> <target dev='hdb' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='0' unit='1'/> </disk> <disk type='file' device='floppy'> <driver name='qemu' type='raw'/> <source file='/var/lib/libvirt/images/1.img'/> <target dev='fda' bus='fdc'/> <address type='drive' controller='0' bus='0' unit='0'/> </disk> <controller type='ide' index='0'/> <controller type='fdc' index='0'/> <interface type='network'> <mac address='52:54:00:5d:46:22'/> <source network='default'/> <target dev='vnet0'/> </interface> <serial type='pty'> <target port='0'/> </serial> <console type='pty'> <target port='0'/> </console> <input type='tablet' bus='usb'/> <input type='mouse' bus='ps2'/> <graphics type='vnc' port='-1' autoport='yes'/> <sound model='es1370'/> <video> <model type='vga' vram='9216' heads='1'/> </video> </devices> </domain> The generated cmdline is: LC_ALL=C PATH=/sbin:/usr/sbin:/bin:/usr/bin QEMU_AUDIO_DRV=none /usr/bin/qemu-kvm -S -M pc-0.11 -enable-kvm -m 1024 -smp 1,sockets=1,cores=1,threads=1 -name winxp_32 -uuid 3485e26d-c370-c0b0-32e2-a041a12f5cfc -nodefaults -chardev socket,id=monitor,path=/var/lib/libvirt/qemu/winxp_32.monitor,server,nowait -mon chardev=monitor,mode=readline -localtime -boot c -drive file=/mnt/devel/vms/disks/winxp_32.qcow2.img,if=none,id=drive-ide0-0-0,boot=on -device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -drive file=/mnt/devel/media/win_2003_64.iso,if=none,media=cdrom,id=drive-ide0-1-0 -device ide-drive,bus=ide.1,unit=0,drive=drive-ide0-1-0,id=ide0-1-0 -drive if=none,media=cdrom,id=drive-ide0-0-1 -device ide-drive,bus=ide.0,unit=1,drive=drive-ide0-0-1,id=ide0-0-1 -drive file=/var/lib/libvirt/images/1.img,if=none,id=drive-fdc0-0-0,format=raw -global isa-fdc,driveA=drive-fdc0-0-0 -device rtl8139,vlan=0,id=net0,mac=52:54:00:5d:46:22,bus=pci.0,addr=0x4 -net tap,fd=38,vlan=0,name=hostnet0 -chardev pty,id=serial0 -device isa-serial,chardev=serial0 -usb -device usb-tablet,id=input0 -vnc 127.0.0.1:0 -vga std -device ES1370,id=sound0,bus=pci.0,addr=0x5 -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 can't parse: "isa-fdc,driveA=drive-fdc0-0-0" If I drop the floppy dev and controller, the guest boots, but a few warnings are printed. Not sure if any of this reflects qemu bugs or libvirt bugs: LC_ALL=C PATH=/sbin:/usr/sbin:/bin:/usr/bin QEMU_AUDIO_DRV=none /usr/bin/qemu-kvm -S -M pc-0.11 -enable-kvm -m 1024 -smp 1,sockets=1,cores=1,threads=1 -name winxp_32 -uuid 3485e26d-c370-c0b0-32e2-a041a12f5cfc -nodefaults -chardev socket,id=monitor,path=/var/lib/libvirt/qemu/winxp_32.monitor,server,nowait -mon chardev=monitor,mode=readline -localtime -boot c -drive file=/mnt/devel/vms/disks/winxp_32.qcow2.img,if=none,id=drive-ide0-0-0,boot=on -device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -drive file=/mnt/devel/media/win_2003_64.iso,if=none,media=cdrom,id=drive-ide0-1-0 -device ide-drive,bus=ide.1,unit=0,drive=drive-ide0-1-0,id=ide0-1-0 -drive if=none,media=cdrom,id=drive-ide0-0-1 -device ide-drive,bus=ide.0,unit=1,drive=drive-ide0-0-1,id=ide0-0-1 -device rtl8139,vlan=0,id=net0,mac=52:54:00:5d:46:22,bus=pci.0,addr=0x4 -net tap,fd=37,vlan=0,name=hostnet0 -chardev pty,id=serial0 -device isa-serial,chardev=serial0 -usb -device usb-tablet,id=input0 -vnc 127.0.0.1:0 -vga std -device ES1370,id=sound0,bus=pci.0,addr=0x5 -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 char device redirected to /dev/pts/2 Warning: vlan 0 with no nics Option 'ipv4': Use 'on' or 'off' Failed to parse "yes" for "dummy.ipv4" Thanks, Cole

2 1

[libvirt] [PATCH 0/9] Remove conn parameter from several functions
by Matthias Bolte 09 Feb '10

09 Feb '10

This set of patches removes the conn parameter from several internal functions where it was used of error reporting only. Since errors are reported to an error object in thread local storage the conn parameter is not necessary for this anymore. Matthias

2 11

[libvirt] [PATCH 2/5] macvtap support for libvirt -- parse new interface XML
by Stefan Berger 08 Feb '10

08 Feb '10

This part adds support to domain_conf.{c|h} for parsing the new interface XML of type 'direct'. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

2 3

[libvirt] [PATCH 4/5] macvtap support for libvirt -- helper code
by Stefan Berger 08 Feb '10

08 Feb '10

This part adds the helper code to setup and tear down macvtap devices using direct communication with the device driver via netlink sockets. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

2 2

[libvirt] RFC: solving problem with qemu domain save on root-squashing NFS
by Laine Stump 08 Feb '10

08 Feb '10

(I had sent this message last week as a response to a message in a patch thread, but thought today that possibly some mail readers would bury it, so I'm sending it again disconnected from any existing thread. I can send the patches I mention below to anyone interested in playing around with them, but they're not useful since they still fail at the final step...) On 01/24/2010 11:57 AM, Dan Kenigsberg wrote: > On Wed, Jan 20, 2010 at 03:14:57PM +0000, Daniel P. Berrange wrote: >> This patch series does some work on te security drivers, and the QEMU >> code >> for managing DAC permissions on files. >> >> The core goal is to turn the QEMU driver DAC file management code into a >> security driver. Instead of QEMU calling into the SELinux/AppArmour >> drivers >> directly, a stacked driver module is introduced. This delegates all >> operations >> to first the QEMU DAC driver, and then the main SELinux/AppArmour >> driver. >> The end result is that all the permissions management code is removed >> from >> the QEMU driver, and we're left with just simple security driver calls. >> >> In the process of this a number of flaws in the current hotplug code >> were >> found, and code was generally tidied up with a view to making it >> easier to >> manage. >> >> Finally, we add the ability to turn off the QEMU DAC file managment >> code, >> and also deal gracefully with failures to change ownership (eg on NFS >> with >> root squash, or readonly FS). > Thanks for this series. However, it seems that we still have a problem > when trying to save domain to a root-squashing nfs export. > When using qemu directly, as a user with write permissions to that > export, there is no problem. When using libvirt, libvirt tries to write > its own state to the target file. I would not want to pre-create the > target file as world redable. > > How about performing open(path, O_CREAT|O_TRUNC|O_WRONLY, > S_IRUSR|S_IWUSR)) with the euid of the qemu process? I came up with a patch that creates the domain state file by first forking, followed by the child process doing setuid(qemu) and all the write-to-the-file stuff that creates the file (similar to the recent virFileCreate patch). This works all the way up through qemu's appending its state onto the file. The trouble comes when qemudDomainSave calls securityDriver->domainRestoreSavedStateLabel() (in this case, qemuSecurityStackedRestoreSavedStateLabel()). That function first calls the primary security driver's qemuSecurityDACRestoreSavedStateLabel(), which will fail because it attempts to chown the file to 0,0 (to prevent other qemu instances from messing with it). That can be disabled by setting "dynamic_ownership=0" in qemu.conf (needed in several other places when stuff is stored on root-squashing NFS anyway :-/). Then the stacked security driver calls the secondary security driver's SELinuxRestoreSavedStateLabel(), which also fails because of an SELinux function (matchpathcon) returning failure. We could ignore this failure, but that would create an unnecessary security hole for all those installations that *aren't* storing their domains on root-squashing NFS. So I *think* what I'm looking for a good way to avoid the SELinux stuff when the destination file is located on a root-squashing NFS share, but not ignore it otherwise (I believe the consensus is that we *don't* want to require another configuration knob if at all possible, since that would not only lead to extra complexity of config, but also to uninformed users turning it off when not necessary). Does anyone have any good ideas for doing this (or for a different approach)? (I suppose I could always attempt the entire operation assuming not, then redo with fork/setuid if the first attempt failed, but that seems less than elegant).

1 0