July 2012 - Devel - libvirt List Archives

[libvirt] Tracking VM installations
by Zeeshan Ali (Khattak) 05 Jul '12

05 Jul '12

Hi, We have a requirement in Boxes for showing progress bars on VMs under installation: https://github.com/gnome-design-team/gnome-mockups/raw/master/boxes/boxes-i… https://bugzilla.gnome.org/show_bug.cgi?id=679107 However, we are bit clueless on how that could be achieved. The only idea I have is to know the number of passes (reboots) and typical disk usage in each pass for "every" (we only do this for express installs so we don't need this for all) OS and estimate the progress based on that knowledge. If that idea is worth looking into, we'll need this information in libosinfo and hence me cross-posting this to virt-tools-list. Other ideas welcome! Keep in mind that we don't need to be very accurate, just accurate enough to not annoy the user. -- Regards, Zeeshan Ali (Khattak) FSF member#5124

3 5

[libvirt] [RFC] Allowing promiscuous mode for domains network interfaces
by Jean-Baptiste Rouault 05 Jul '12

05 Jul '12

Hi all, By default, OpenVZ and VirtualBox (> 4.0.x) filter network packets by MAC addresses : only broadcast, multicast and packets directly targeted to VMs are transmitted. This behaviour prevents from using promiscuous mode inside domains. I'd like to write some patches to disable these filters from libvirt. Would it be ok to modify OpenVZ and VirtualBox drivers so that they disable the filters by default ? If this is not acceptable, what about making it configurable through domains' XML ? Regards, Jean-Baptiste -- Jean-Baptiste ROUAULT Ingénieur R&D - diateam : Architectes de l'information Phone : +33 (0)2 98 050 050 Fax : +33 (0)2 98 050 051

2 2

[libvirt] [PATCH v7 0/8] Add basic driver for Parallels Virtuozzo Server
by Dmitry Guryanov 04 Jul '12

04 Jul '12

Parallels Cloud Server is a virtualization solution that allows users to simultaneously run multiple virtual machines and containers on the same physical server. More information can be found here: http://www.parallels.com/products/pcs/ Also beta version of Parallels Cloud Server can be downloaded there. Dmitry Guryanov (8): parallels: add driver skeleton parallels: add functions to list domains and get info parallels: implement functions for domain life cycle management parallels: get info about serial ports parallels: add support of VNC remote display parallels: implement virDomainDefineXML operation for existing domains parallels: add storage driver parallels: implement VM creation cfg.mk | 1 + configure.ac | 23 + docs/drvparallels.html.in | 28 + include/libvirt/virterror.h | 2 +- libvirt.spec.in | 9 +- mingw32-libvirt.spec.in | 6 + po/POTFILES.in | 4 + src/Makefile.am | 15 + src/conf/domain_conf.c | 3 +- src/conf/domain_conf.h | 1 + src/driver.h | 1 + src/libvirt.c | 9 + src/parallels/parallels_driver.c | 1309 +++++++++++++++++++++++++++++++++ src/parallels/parallels_driver.h | 75 ++ src/parallels/parallels_storage.c | 1456 +++++++++++++++++++++++++++++++++++++ src/parallels/parallels_utils.c | 131 ++++ src/util/virterror.c | 3 +- 17 files changed, 3072 insertions(+), 4 deletions(-) create mode 100644 docs/drvparallels.html.in create mode 100644 src/parallels/parallels_driver.c create mode 100644 src/parallels/parallels_driver.h create mode 100644 src/parallels/parallels_storage.c create mode 100644 src/parallels/parallels_utils.c

3 16

[libvirt] can't receive keepalive message during migration
by Hu Tao 04 Jul '12

04 Jul '12

Hi, When testing migration, I found a problem that the server can't receive keepalive reply from client when doing migration. The log shows that virsh did send the keepalive reply, but server didn't reveive it. The strange part is it's OK to migrate from host A to Host B, but always fails to migrate from host B to host A. When disable keepalive, migration will success. Any ideas? -- Thanks, Hu Tao

2 2

[libvirt] [PATCH] virsh: Avoid division by 0 in vshCalloc
by Peter Krempa 04 Jul '12

04 Jul '12

vshCalloc function uses xalloc_oversized macro that can't take 0 as it's second argument. If vshCalloc is called with size 0, virsh ends with a floating point exception. This patch changes vshCalloc to return NULL if no memory is requested. --- tools/virsh.c | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 53d1825..d3d5c6a 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -460,6 +460,9 @@ _vshCalloc(vshControl *ctl, size_t nmemb, size_t size, const char *filename, int { char *x; + if (!size) + return NULL; + if (!xalloc_oversized(nmemb, size) && VIR_ALLOC_N(x, nmemb * size) == 0) return x; -- 1.7.8.6

3 5

[libvirt] [PATCH] Only enforce check for YAJL when starting a VM
by Daniel P. Berrange 04 Jul '12

04 Jul '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> The previous check for YAJL would have many undesirable consequences, the most important being that it caused the capabilities XML to loose all <guest> elements. There is no user visible feedback as to what is wrong in this respect, merely a syslog message. The empty capabilities causes libvirtd to then throw away all guest XML configs that are stored. This changes the code so that the check for YAJL is only performed at the time we attempt to spawn a QEMU process error: Failed to start domain vm-vnc error: unsupported configuration: this qemu binary requires libvirt to be compiled with yajl Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- src/qemu/qemu_capabilities.c | 13 +++++++++---- src/qemu/qemu_capabilities.h | 4 +++- src/qemu/qemu_command.c | 1 + src/qemu/qemu_driver.c | 1 + src/qemu/qemu_process.c | 3 +++ 5 files changed, 17 insertions(+), 5 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 1e12a39..1d6ab2d 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -691,7 +691,8 @@ qemuCapsInitGuest(virCapsPtr caps, return 0; /* Ignore binary if extracting version info fails */ - if (qemuCapsExtractVersionInfo(binary, info->arch, NULL, &qemuCaps) < 0) { + if (qemuCapsExtractVersionInfo(binary, info->arch, + false, NULL, &qemuCaps) < 0) { ret = 0; goto cleanup; } @@ -1472,7 +1473,9 @@ qemuCapsParseDeviceStr(const char *str, virBitmapPtr flags) return 0; } -int qemuCapsExtractVersionInfo(const char *qemu, const char *arch, +int qemuCapsExtractVersionInfo(const char *qemu, + const char *arch, + bool check_yajl, unsigned int *retversion, virBitmapPtr *retflags) { @@ -1505,7 +1508,8 @@ int qemuCapsExtractVersionInfo(const char *qemu, const char *arch, if (!(flags = qemuCapsNew()) || qemuCapsParseHelpStr(qemu, help, flags, - &version, &is_kvm, &kvm_version, true) == -1) + &version, &is_kvm, &kvm_version, + check_yajl) == -1) goto cleanup; /* Currently only x86_64 and i686 support PCI-multibus. */ @@ -1583,7 +1587,8 @@ int qemuCapsExtractVersion(virCapsPtr caps, return -1; } - if (qemuCapsExtractVersionInfo(binary, ut.machine, version, NULL) < 0) { + if (qemuCapsExtractVersionInfo(binary, ut.machine, false, + version, NULL) < 0) { return -1; } diff --git a/src/qemu/qemu_capabilities.h b/src/qemu/qemu_capabilities.h index 83c135b..c14d53a 100644 --- a/src/qemu/qemu_capabilities.h +++ b/src/qemu/qemu_capabilities.h @@ -167,7 +167,9 @@ int qemuCapsProbeCPUModels(const char *qemu, int qemuCapsExtractVersion(virCapsPtr caps, unsigned int *version); -int qemuCapsExtractVersionInfo(const char *qemu, const char *arch, +int qemuCapsExtractVersionInfo(const char *qemu, + const char *arch, + bool check_yajl, unsigned int *version, virBitmapPtr *qemuCaps); diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 528b189..5959352 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -949,6 +949,7 @@ qemuDomainAssignPCIAddresses(virDomainDefPtr def) qemuDomainPCIAddressSetPtr addrs = NULL; if (qemuCapsExtractVersionInfo(def->emulator, def->os.arch, + false, NULL, &qemuCaps) < 0) goto cleanup; diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 2f93404..eb2ed0e 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -4757,6 +4757,7 @@ static char *qemuDomainXMLToNative(virConnectPtr conn, } if (qemuCapsExtractVersionInfo(def->emulator, def->os.arch, + false, NULL, &qemuCaps) < 0) goto cleanup; diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index c5140c3..fd4c9c9 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -3073,6 +3073,7 @@ qemuProcessReconnect(void *opaque) */ if (!priv->qemuCaps && qemuCapsExtractVersionInfo(obj->def->emulator, obj->def->os.arch, + false, NULL, &priv->qemuCaps) < 0) goto error; @@ -3478,6 +3479,7 @@ int qemuProcessStart(virConnectPtr conn, qemuCapsFree(priv->qemuCaps); priv->qemuCaps = NULL; if (qemuCapsExtractVersionInfo(vm->def->emulator, vm->def->os.arch, + true, NULL, &priv->qemuCaps) < 0) goto cleanup; @@ -4237,6 +4239,7 @@ int qemuProcessAttach(virConnectPtr conn ATTRIBUTE_UNUSED, priv->qemuCaps = NULL; if (qemuCapsExtractVersionInfo(vm->def->emulator, vm->def->os.arch, + false, NULL, &priv->qemuCaps) < 0) goto cleanup; -- 1.7.10.4

2 1

[libvirt] [PATCH] virsh: Fix ordering of arguments when calling vshCalloc.
by Peter Krempa 04 Jul '12

04 Jul '12

In vshSnapshotListCollect() vshCalloc was called with swapped nmemb and size argument. This caused division by zero in xalloc_oversized as the macro doesn't expect size to be zero. --- tools/virsh.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 53d1825..25ed177 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -17200,8 +17200,8 @@ vshSnapshotListCollect(vshControl *ctl, virDomainPtr dom, if (count >= 0) { /* When mixing --from and --tree, we also want a copy of from * in the list, but with no parent for that one entry. */ - snaplist->snaps = vshCalloc(ctl, sizeof(*snaplist->snaps), - count + (tree && from)); + snaplist->snaps = vshCalloc(ctl, pcount + (tree && from), + sizeof(*snaplist->snaps)); snaplist->nsnaps = count; for (i = 0; i < count; i++) snaplist->snaps[i].snap = snaps[i]; -- 1.7.8.6

2 3

[libvirt] [PATCH] qemu: Mark domains as having managed state image only on managed save
by Peter Krempa 04 Jul '12

04 Jul '12

QEmu domains were marked as having managed save image even if they were saved using the regular save. With this patch, domains are marked so only when using managed save API. --- src/qemu/qemu_driver.c | 7 ++++--- 1 files changed, 4 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 2f93404..3410535 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2739,7 +2739,6 @@ qemuDomainSaveInternal(struct qemud_driver *driver, virDomainPtr dom, } ret = 0; - vm->hasManagedSave = true; /* Shut it down */ qemuProcessStop(driver, vm, VIR_DOMAIN_SHUTOFF_SAVED, 0); @@ -2916,8 +2915,10 @@ qemuDomainManagedSave(virDomainPtr dom, unsigned int flags) VIR_INFO("Saving state to %s", name); compressed = QEMUD_SAVE_FORMAT_RAW; - ret = qemuDomainSaveInternal(driver, dom, vm, name, compressed, - NULL, flags); + if ((ret = qemuDomainSaveInternal(driver, dom, vm, name, compressed, + NULL, flags)) == 0) + vm->hasManagedSave = true; + vm = NULL; cleanup: -- 1.7.8.6

3 3

[libvirt] question about save/restore domain
by Jovanka Gulicoska 04 Jul '12

04 Jul '12

Hi, my name is Jovanka Gulicoska and I'm SoC student. I'm working on implementing save/load of VM in Gnome-Boxes. Something confused me about snapshots. In the documentation about life cycle, in section about Save/Restore is said that domain saved with virDomainSave/virDomainSaveFlags can be restored only once. But in the documentation about Snapshot XML format there is no information about this in VM state snapshots. So is it better just to use virDomainSnapshotCreateXML instead of virDomainSave/virDomainSaveFlags? I've created bindings for this functions in libvirt-glib but I wanted to be sure before I send them for review. Thanks, Jovanka (gulic)

5 8

[libvirt] libvirt can't get capablities
by Li Zhang 04 Jul '12

04 Jul '12

Hi, I have suffered several times on both x86 machine and power machine. When we want to use libvirt to create on VMs, it reports "unknown OS type hvm". We can't get <guest> capabilities. The log is as the following(log_level=1): 2003-01-02 01:21:39.895+0000: 12214: error : virDomainDefParseXML:8491 : unknown OS type hvm 2003-01-02 01:21:39.896+0000: 12214: error : virDomainDefParseXML:8491 : unknown OS type hvm 2003-01-02 01:22:00.135+0000: 12207: error : qemuCapsComputeCmdFlags:1218 : unsupported configuration: this qemu binary requires libvirt to be compiled with yajl 2003-01-02 01:22:13.342+0000: 12207: error : qemuCapsComputeCmdFlags:1218 : unsupported configuration: this qemu binary requires libvirt to be compiled with yajl 2003-01-02 01:22:13.347+0000: 12204: error : qemuCapsExtractVersion:1576 : internal error Cannot find suitable emulator for ppc64 2003-01-02 01:22:13.351+0000: 12205: error : qemuCapsExtractVersion:1576 : internal error Cannot find suitable emulator for ppc64 2003-01-02 01:22:13.355+0000: 12207: error : qemuCapsExtractVersion:1576 : internal error Cannot find suitable emulator for ppc64 2003-01-02 01:22:13.355+0000: 12208: error : qemuCapsExtractVersion:1576 : internal error Cannot find suitable emulator for ppc64 2003-01-02 01:22:13.357+0000: 12206: error : qemuCapsExtractVersion:1576 : internal error Cannot find suitable emulator for ppc64 I search online and find one similar issue from Redhat bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=509337 Is there any solution to avoid this problem? Any idea? -- Best Regards Li IBM LTC, China System&Technology Lab, Beijing

4 10