- Devel - libvirt List Archives

[libvirt] [PATCH] util: Fix the indention
by Osier Yang 25 Nov '12

25 Nov '12

--- Pushed under trivial rule. --- src/util/cgroup.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/src/util/cgroup.c b/src/util/cgroup.c index 3f7b5f7..41a6e4a 100644 --- a/src/util/cgroup.c +++ b/src/util/cgroup.c @@ -1189,10 +1189,10 @@ int virCgroupSetBlkioDeviceWeight(virCgroupPtr group, return ret; } #else - int - virCgroupSetBlkioDeviceWeight(virCgroupPtr group ATTRIBUTE_UNUSED, - const char *path ATTRIBUTE_UNUSED, - unsigned int weight ATTRIBUTE_UNUSED) +int +virCgroupSetBlkioDeviceWeight(virCgroupPtr group ATTRIBUTE_UNUSED, + const char *path ATTRIBUTE_UNUSED, + unsigned int weight ATTRIBUTE_UNUSED) { return -ENOSYS; } -- 1.7.7.6

1 0

[libvirt] image transfer incompletely in live migration
by yli＠cloudiya.com 25 Nov '12

25 Nov '12

Hi, I am using libvirt version 0.10.1 and qemu-kvm version 1.2.0 with RHEL 6.3 When I use libvirt API to make block migrate with a VM, it failed frequently. I check the migrated VM and found that the filesystem damaged. But sometimes it works very fine, I don't know why, it seems to be random. All failed tests have the same condition, image transfer incompletely in live migration. The image I used is 10GB size, after a failed migration it has only about 400MB to 3GB in remote host. Here is the code: import libvirt from xml.dom import minidom name = "vm-01" host = libvirt.open("qemu:///system") vm = host.lookupByNmae(name) vm_xml_file = "/etc/libvirt/qemu/%s.xml" % name vm_xml = minidom.parse(vm_xml_file).documentElement.toxml() remote_host = "qemu+ssh://host02/system" remote = libvirt.open(remote_host) try: vm.migrate2(remote, vm_xml, 89, name, None, 0) except Exception,e: print str(e) Before run the code I have already create a image for migration in remote host 'host02'. Path, size, and privilege all OK. Here is the related log: in source host 'host01': 2012-11-24 05:14:40.001+0000: 24888: warning : qemuDomainObjBeginJobInternal:838 : Cannot start job (query, none) for domain wrksp-776; current job is (async nested, migration out) owned by (24890, 24890) 2012-11-24 05:16:16.398+0000: 24888: error : qemuDomainObjBeginJobInternal:842 : Timed out during operation: cannot acquire state change lock 2012-11-24 05:16:21.317+0000: 24887: error : virNetSocketReadWire:1184 : End of file while reading data: Input/output error in remote host 'host02': 2012-11-24 05:14:00.776+0000: 11065: warning : qemuDomainObjEnterMonitorInternal:993 : This thread seems to be the async job owner; entering monitor without asking for a nested job is dangerous 2012-11-24 05:14:01.196+0000: 11065: warning : qemuDomainObjEnterMonitorInternal:993 : This thread seems to be the async job owner; entering monitor without asking for a nested job is dangerous 2012-11-24 05:14:01.198+0000: 11065: warning : qemuDomainObjEnterMonitorInternal:993 : This thread seems to be the async job owner; entering monitor without asking for a nested job is dangerous 2012-11-24 05:14:01.208+0000: 11065: warning : qemuDomainObjEnterMonitorInternal:993 : This thread seems to be the async job owner; entering monitor without asking for a nested job is dangerous 2012-11-24 05:14:01.211+0000: 11065: warning : qemuDomainObjEnterMonitorInternal:993 : This thread seems to be the async job owner; entering monitor without asking for a nested job is dangerous I try to use command "virsh migrate vm-01 --live --copy-storage-all --persistent --undefinesource qemu+ssh://host02/system --verbose", the test results have no change, common but not regular. Even after my once successful test, it will certainly fail when I try to make a second migrate with the same VM. How can I solve this problem？ Anything is helpful. Thanks.

1 0

[libvirt] [PATCH 0/6] Misc unrelated fixes
by Daniel P. Berrange 25 Nov '12

25 Nov '12

The following are miscellaneous patches fixes various issues I've come across today, while hacking on LXC

4 12

[libvirt] [BUG] storage.xml: owner|group=-1 → 2^32-1 on 32 Bit
by Philipp Hahn 23 Nov '12

23 Nov '12

Hello, when you install and use libvirt in a 32 bit environment, the handling of owner and group in /etc/libvirt/storage/*.xml is broken: The initial -1 is printed as an "unsigned int" and is thus converted to +(2^32-1): > # cat /etc/libvirt/storage/default.xml >  > > <pool type='dir'> > <name>default</name> > <uuid>4e2670b8-9dc3-528e-6fc2-58268e99f44f</uuid> > <capacity unit='bytes'>0</capacity> > <allocation unit='bytes'>0</allocation> > <available unit='bytes'>0</available> > <source> > </source> > <target> > <path>/var/lib/libvirt/images</path> > <permissions> > <mode>0700</mode> > <owner>4294967295</owner> > <group>4294967295</group> > </permissions> > </target> > </pool> When you next restart libvirtd virStorageDefParsePerms() tries to parse that as an "signed int", which fails in virXPathLongBase() because of > 182 if (*value != obj->floatval) { > (gdb) print *value > $11 = -2147483648 > (gdb) print obj->floatval > $12 = 4294967295 > (gdb) print /x *value > $13 = 0x80000000 > (gdb) print /x obj->floatval > $14 = 0xffffffff This then prints the following error messages: > 2012-11-23 14:03:18.877+0000: 25536: error : virStorageDefParsePerms:613 : > XML error: malformed owner element > 2012-11-23 14:03:19.581+0000: > 25529: error : storagePoolSetAutostart:1056 : Failed to create symlink > '/etc/libvirt/storage/autostart/default.xml' to '/etc/libvirt/storage/def The change to print owner and group as %d instead of %u was introdeces in commit 37a10129: > Update xml schemas according to libvirt source ... > storage_conf: Print uid_t and gid_t as signed to storage pool XML. Care to revert that change? Patch is attached. Sincerely Philipp -- Philipp Hahn Open Source Software Engineer hahn(a)univention.de Univention GmbH be open. fon: +49 421 22 232- 0 Mary-Somerville-Str.1 D-28359 Bremen fax: +49 421 22 232-99 http://www.univention.de/ diff --git a/src/conf/storage_conf.c b/src/conf/storage_conf.c index 7944555..b07a7aa 100644 --- a/src/conf/storage_conf.c +++ b/src/conf/storage_conf.c @@ -1036,10 +1036,10 @@ virStoragePoolDefFormat(virStoragePoolDefPtr def) { virBufferAddLit(&buf," <permissions>\n"); virBufferAsprintf(&buf," <mode>0%o</mode>\n", def->target.perms.mode); - virBufferAsprintf(&buf," <owner>%d</owner>\n", - (int) def->target.perms.uid); - virBufferAsprintf(&buf," <group>%d</group>\n", - (int) def->target.perms.gid); + virBufferAsprintf(&buf," <owner>%u</owner>\n", + (unsigned int) def->target.perms.uid); + virBufferAsprintf(&buf," <group>%u</group>\n", + (unsigned int) def->target.perms.gid);

2 1

[libvirt] [PATCHv2] virsh: Report error when taking a snapshot with empty --memspec argument
by Peter Krempa 23 Nov '12

23 Nov '12

When the value of memspec was empty taking of a snapshot failed without reporting an error. --- tools/virsh-snapshot.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/tools/virsh-snapshot.c b/tools/virsh-snapshot.c index 398730c..8ec6456 100644 --- a/tools/virsh-snapshot.c +++ b/tools/virsh-snapshot.c @@ -358,18 +358,19 @@ cmdSnapshotCreateAs(vshControl *ctl, const vshCmd *cmd) if (desc) virBufferEscapeString(&buf, " <description>%s</description>\n", desc); - if (vshCommandOptString(cmd, "memspec", &memspec) < 0 || - vshParseSnapshotMemspec(ctl, &buf, memspec) < 0) { - virBufferFreeAndReset(&buf); + if (vshCommandOptString(cmd, "memspec", &memspec) < 0) { + vshError(ctl, _("memspec argument must not be empty")); goto cleanup; } + + if (memspec && vshParseSnapshotMemspec(ctl, &buf, memspec) < 0) + goto cleanup; + if (vshCommandOptBool(cmd, "diskspec")) { virBufferAddLit(&buf, " <disks>\n"); while ((opt = vshCommandOptArgv(cmd, opt))) { - if (vshParseSnapshotDiskspec(ctl, &buf, opt->data) < 0) { - virBufferFreeAndReset(&buf); + if (vshParseSnapshotDiskspec(ctl, &buf, opt->data) < 0) goto cleanup; - } } virBufferAddLit(&buf, " </disks>\n"); } @@ -390,6 +391,7 @@ cmdSnapshotCreateAs(vshControl *ctl, const vshCmd *cmd) ret = vshSnapshotCreate(ctl, dom, buffer, flags, NULL); cleanup: + virBufferFreeAndReset(&buf); VIR_FREE(buffer); if (dom) virDomainFree(dom); -- 1.8.0

2 2

Re: [libvirt] Remote connection to KVM hypervisor failed using libvirt 0.10.2
by Daniel P. Berrange 23 Nov '12

23 Nov '12

On Fri, Nov 23, 2012 at 11:13:58AM +0200, Ido Levy wrote: > > From: "Daniel P. Berrange" <berrange(a)redhat.com> > > To: Ido Levy/Haifa/IBM@IBMIL, > > Cc: libvir-list(a)redhat.com > > Date: 23/11/2012 10:59 AM > > Subject: Re: [libvirt] Remote connection to KVM hypervisor failed > > using libvirt 0.10.2 > > > > On Fri, Nov 23, 2012 at 07:43:12AM +0200, Ido Levy wrote: > > > Hello, > > > > > > I compiled libvirt 0.10.2 from source providing extra parameters to > the > > > configure procedure to support phyp and hyper-v driver support, and > python > > > binding. > > > When running virsh from the tools directory (libvirt-0.10.2/tools) and > > > > trying to connect to remote KVM hypervisor using connect > > > qemu+ssh://user@hostname/system > > > I got the following error > > > error: Failed to connect to the hypervisor > > > error: End of file while reading data: : Input/output error > > > > > > The same scenario working without any issue using virsh tool that is > > > shipped with RHEL distribution. > > > > > > Any advise how can I figure out how to resolve this. > > > > What are the full args you passed to 'configure' ? > > > > Daniel > > ./configure --with-python --with-phyp --with-hyperv Ok, that's the problem. If you want your build to be able to connect to the RPM based build, you need to make it use the same paths. For this you need to pass ./configure --with-python --with-phyp --with-hyperv \ --prefix=/usr --sysconfdir=/etc --localstatedir=/var Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|

2 2

[libvirt] [PATCH 0/4] qemu: add boot order support for redirected and host USB devices
by Ján Tomko 23 Nov '12

23 Nov '12

Commit a4c19459aa8634c43b51e8138fb1d7eec4c17824 missed the functionality for redirected USB devices (fixed in 1/4), documentation and tests. https://bugzilla.redhat.com/show_bug.cgi?id=805414 Ján Tomko (4): conf: add support for booting from redirected USB devices docs: boot order for host and redirected USB devices tests: add boot order for host and redirected USB to qemu argv test tests: update qemuhelptest data docs/formatdomain.html.in | 27 ++- src/conf/domain_conf.c | 34 ++-- tests/qemuhelpdata/qemu-1.2.0 | 1 + tests/qemuhelpdata/qemu-1.2.0-device | 51 +++- tests/qemuhelpdata/qemu-kvm-1.2.0 | 276 ++++++++++++++++++++ tests/qemuhelpdata/qemu-kvm-1.2.0-device | 222 ++++++++++++++++ tests/qemuhelptest.c | 106 ++++++++- ...muxml2argv-hostdev-usb-address-device-boot.args | 6 + ...emuxml2argv-hostdev-usb-address-device-boot.xml | 28 ++ .../qemuxml2argv-usb-redir-boot.args | 10 + .../qemuxml2argv-usb-redir-boot.xml | 42 +++ tests/qemuxml2argvtest.c | 9 + 12 files changed, 775 insertions(+), 37 deletions(-) create mode 100644 tests/qemuhelpdata/qemu-kvm-1.2.0 create mode 100644 tests/qemuhelpdata/qemu-kvm-1.2.0-device create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdev-usb-address-device-boot.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdev-usb-address-device-boot.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-usb-redir-boot.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-usb-redir-boot.xml -- 1.7.8.6

3 13

[libvirt] [PATCH] bitmap: fix typo to use UL type of integer constant in virBitmapIsAllSet
by Guannan Ren 23 Nov '12

23 Nov '12

This bug leads to get incorrect vcpupin information via qemudDomainGetVcpuPinInfo() API when the number of maximum cpu on a host falls into a scope such as 31 < ncpus < 64. gcc warning: left shift count >= width of type The following bug is such a case https://bugzilla.redhat.com/show_bug.cgi?id=876415 --- src/util/bitmap.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/util/bitmap.c b/src/util/bitmap.c index 5ec5440..c29f5f3 100644 --- a/src/util/bitmap.c +++ b/src/util/bitmap.c @@ -574,8 +574,8 @@ bool virBitmapIsAllSet(virBitmapPtr bitmap) return false; if (unusedBits > 0) { - if ((bitmap->map[sz] & ((1U << (VIR_BITMAP_BITS_PER_UNIT - unusedBits)) - 1)) - != ((1U << (VIR_BITMAP_BITS_PER_UNIT - unusedBits)) - 1)) + if ((bitmap->map[sz] & ((1UL << (VIR_BITMAP_BITS_PER_UNIT - unusedBits)) - 1)) + != ((1UL << (VIR_BITMAP_BITS_PER_UNIT - unusedBits)) - 1)) return false; } -- 1.7.1

2 2

[libvirt] Remote connection to KVM hypervisor failed using libvirt 0.10.2
by Ido Levy 23 Nov '12

23 Nov '12

Hello, I compiled libvirt 0.10.2 from source providing extra parameters to the configure procedure to support phyp and hyper-v driver support, and python binding. When running virsh from the tools directory (libvirt-0.10.2/tools) and trying to connect to remote KVM hypervisor using connect qemu+ssh://user@hostname/system I got the following error error: Failed to connect to the hypervisor error: End of file while reading data: : Input/output error The same scenario working without any issue using virsh tool that is shipped with RHEL distribution. Any advise how can I figure out how to resolve this. Thanks, Ido

3 2

[libvirt] QMP Capability Detection Issues with GIT version of QEMU
by Viktor Mihajlovski 23 Nov '12

23 Nov '12

I ran into trouble today trying to use the GIT level of QEMU. In a nutshell: the capability detection with QMP is failing and the fallback using -help isn't working with the GIT level probably due to help text reformatting. The failure reason is that QEMU cannot bind to the QMP monitor socket in the /var/lib/libvirt/qemu directory. That's because the child process is stripped of all capabilities and this directory is chown'ed to qemu:qemu by the QEMU driver. Note that this is failing with the release QEMU as well, with the difference that the fallback is working there. I am willing to provide a patch, however I'd like to get feedback on the approach to use: 1. Add back Linux capabilities CAP_DAC_OVERRIDE, CAP_DAC_READ_SEARCH to allow QEMU to bind to the monitor socket. This seems to be hacky/dodging the problem. 2. Use a separate directory for the QMP probing instance of QEMU. 3. Run the QMP QEMU under the configured qemu user. This would be my favorite. 4. Other ideas? -- Mit freundlichen Grüßen/Kind Regards Viktor Mihajlovski IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294

2 3