Re: [libvirt] [PATCH v5 3/3] gluster: cache glfs connection object per volume

On Thu, Dec 15, 2016 at 13:37:05 +0530, Prasanna Kumar Kalever wrote: > Currently, in case if we have 4 extra attached disks, then for each disk > we need to call 'glfs_init' (over network) and friends which could be costly. > > Additionally snapshot(external) scenario will further complex the

Re: [libvirt] [PATCH v5 0/3] gluster: cache glfs connection object per volume

Hi Virt Team, Please review this patch series. Thanks, -- Prasanna On Thu, Dec 15, 2016 at 1:37 PM, Prasanna Kumar Kalever wrote: > v5: Address review comments from Peter on v4 > Dropping Patch 3 from v3 > Patch 1: fix transport type in all drivers, possible

[libvirt] [PATCH 0/9] i386: query-cpu-model-expansion test script

This is a follow-up to the series that implements query-cpu-model-expansion. Before including the test script, the series has some fixes to allow the results of query-cpu-model-expansion to be used in the QEMU command-line. The script probably will work on s390x too, but I couldn't test it yet.

[libvirt] [PATCH v2 0/4] target-i386: Implement query-cpu-model-expansion

This series implements query-cpu-model-expansion on target-i386. Changes v1 -> v2: - This version is highly simplified compared to v1. It contains only an implementation that will return a limited set of properties. I have a follow-up series that will expend type=full expansion

[libvirt] [PATCH v2 4/4] target-i386: Implement query-cpu-model-expansion QMP command

Implement query-cpu-model-expansion for target-i386. This should meet all the requirements while being simple. In the case of static expansion, it will use the new "base" CPU model, and in the case of full expansion, it will keep the original CPU model name+props, and append extra properties. A

Re: [libvirt] Performance about x-data-plane

On Mon, Jan 16, 2017 at 8:15 AM, Stefan Hajnoczi wrote: > On Tue, Jan 03, 2017 at 12:02:14PM -0500, Weiwei Jia wrote: >> > The expensive part is the virtqueue kick. Recently we tried polling the >> > virtqueue instead of waiting for the ioeventfd file descriptor and got >> >

Re: [libvirt] Performance about x-data-plane

On Mon, Jan 16, 2017 at 10:00 AM, Karl Rister wrote: > On 01/16/2017 07:15 AM, Stefan Hajnoczi wrote: >> On Tue, Jan 03, 2017 at 12:02:14PM -0500, Weiwei Jia wrote: The expensive part is the virtqueue kick. Recently we tried polling the virtqueue instead of waiting

[libvirt] [PATCH] tests: fix QED disk test in xlconfigtest

When LIBXL_HAVE_QED is defined, xlconfigtest fails 9) Xen XL-2-XML Format disk-qed ... command line: config parsing error in disk specification: no vdev specified in `target=/var/lib/libvirt/images/XenGuest2,format=qed,backendtype=qdisk,vdev=hda,access=rw' FAILED As per the

Re: [libvirt] Performance about x-data-plane

On 01/16/2017 07:15 AM, Stefan Hajnoczi wrote: > On Tue, Jan 03, 2017 at 12:02:14PM -0500, Weiwei Jia wrote: >>> The expensive part is the virtqueue kick. Recently we tried polling the >>> virtqueue instead of waiting for the ioeventfd file descriptor and got >>> double-digit performance

[libvirt] [PATCH] tests: fix compilation of shunloadtest

While local builds succeed fine, a build worker building in a chroot environment is encountering the following error with libvirt 3.0.0 release candidates [ 162s] shunloadtest.o: In function `main': [ 162s] /home/abuild/rpmbuild/BUILD/libvirt-3.0.0/tests/shunloadtest.c:110: undefined reference

Re: [libvirt] [PATCH for 3.0.0] Revert "perf: Add cache_l1d perf event support"

On Mon, Jan 16, 2017 at 12:07:39PM -0500, John Ferlan wrote: > > > On 01/16/2017 11:57 AM, Daniel P. Berrange wrote: > > This reverts commit ae16c95f1bb5591c27676c5de8d383e5612c3568. > > > > The data was calculated incorrectly and the event name needs > > to be changed. > > > > --- > >

Re: [libvirt] [PATCH for 3.0.0] Revert "perf: Add cache_l1d perf event support"

On 01/16/2017 11:57 AM, Daniel P. Berrange wrote: > This reverts commit ae16c95f1bb5591c27676c5de8d383e5612c3568. > > The data was calculated incorrectly and the event name needs > to be changed. > > --- > docs/formatdomain.html.in | 7 --- >

[libvirt] [PATCH for 3.0.0] Revert "perf: Add cache_l1d perf event support"

This reverts commit ae16c95f1bb5591c27676c5de8d383e5612c3568. The data was calculated incorrectly and the event name needs to be changed. --- docs/formatdomain.html.in | 7 --- docs/schemas/domaincommon.rng | 1 - include/libvirt/libvirt-domain.h

Re: [libvirt] Availability of libvirt-3.0.0 release candidate 2

On Sat, Jan 14, 2017 at 11:02:37PM +0100, Daniel Veillard wrote: > I missed the Friday deadline ot push it, so did it today. It's now tagged > in git, signed tarball and rpms are at the usual place: >ftp://libvirt.org/libvirt/ > > As a result I think GA should happen on Tuesday to give at

[libvirt] vms hung

Hello, I am running libvirt-2.0.0-10.el7_3.2.x86_64 on CentOS 7.3, I have few vms of CentOS and Windows 2012 servers. Randomly they are hung, but the status of virtual machine keep running, vnc and network not responding. I have checked all /var/log/messages and /var/log/libvirt/qemu/vm.log but

[libvirt] [RFE] Enable/allow MAC spoofing for SR-IOV virtual functions.

Hey, We've recently encountered difficulties using a virtual function as a slave in a bond where the virtual function's MAC address differs from the bond's. I've opened an RFE[1] suggesting MAC spoofing to be enabled/allowed in virtual functions. [1]

Re: [libvirt] [PATCH] qemu: Copy SELinux labels for namespace too

On Fri, Jan 13, 2017 at 03:47:16PM +0100, Michal Privoznik wrote: On 01/13/2017 12:43 PM, Martin Kletzander wrote: On Fri, Jan 13, 2017 at 11:12:43AM +0100, Michal Privoznik wrote: When creating new /dev/* for qemu, we do chown() and copy ACLs to create the exact copy from the original /dev. I

Re: [libvirt] [PATCH v2] Add support for Veritas HyperScale (VxHS) block device protocol

On 01/16/2017 04:07 AM, Peter Krempa wrote: > On Sat, Jan 14, 2017 at 10:31:05 -0500, John Ferlan wrote: >> [...] >> diff --git a/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-vxhs.args b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-vxhs.args new file

Re: [libvirt] How to generate better API documentation?

On Mon, Jan 16, 2017 at 09:17:03AM +, Daniel P. Berrange wrote: On Mon, Jan 16, 2017 at 08:24:25AM +0100, Michal Privoznik wrote: Dear list, now that we more or less agreed to use some new features (i.e. automatic free() when a variable goes out of the scope [1]), and with our NEWS

[libvirt] [PATCH v3 2/2] nodedev: Fabric name must not be required for fc_host capability

fabric_name is one of many fc_host attributes in Linux that is optional and left to the low-level driver to decide if it is implemented. The zfcp device driver does not provide a fabric name for an fcp host. This patch removes the requirement for a fabric name by making it optional.

[libvirt] [PATCH v3 1/2] util: add file exists check in virReadFCHost

File open errors are prevented by a file exists check before virFileReadAll is called since all callers of the virReadFCHost method handle errors themselves based on the NULL return anyway. Also included is a minor spelling correction in a comment. Signed-off-by: Boris Fiuczynski

[libvirt] [PATCH v3 0/2] Fabric name must not be required for fc_host capability

fabric_name is one of many fc_host attributes in Linux that is optional and left to the low-level driver to decide if it is implemented. --- Changes sinve v2: - removed virReadFCHostOption method - changed additional file check to be used permanently Changes since v1: - split of minor

Re: [libvirt] Performance about x-data-plane

On Tue, Jan 03, 2017 at 12:02:14PM -0500, Weiwei Jia wrote: > > The expensive part is the virtqueue kick. Recently we tried polling the > > virtqueue instead of waiting for the ioeventfd file descriptor and got > > double-digit performance improvements: > >

[libvirt] [PATCH] qemu_domain: add timestamp in tainting of guests log

From: Chen Hanxiao We lacked of timestamp in tainting of guests log, which bring troubles for finding guest issues: such as whether a guest powerdown caused by qemu-monitor-command or others issues inside guests. If we had timestamp in tainting of guests log, it would be

Re: [libvirt] [PATCH 1/2] perf: Compute cache_l1d config value correctly

On Sat, Jan 14, 2017 at 01:49:59PM +0530, Nitesh Konkar wrote: > This patch computes the .attrConfig value for > cache_l1d correctly and updates the documentation. > The cache_l1d perf event now is renamed as > cache_l1dra perf event for measuring read accesses > for level 1 data cache > >

Re: [libvirt] How to generate better API documentation?

On Mon, Jan 16, 2017 at 08:24:25AM +0100, Michal Privoznik wrote: > Dear list, > > now that we more or less agreed to use some new features (i.e. automatic > free() when a variable goes out of the scope [1]), and with our NEWS efforts, > do you think it is finally the right time to fix

Re: [libvirt] [PATCH v2] Add support for Veritas HyperScale (VxHS) block device protocol

On Sat, Jan 14, 2017 at 10:31:05 -0500, John Ferlan wrote: > [...] > > >> diff --git > >> a/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-vxhs.args > >> b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-vxhs.args > >> new file mode 100644 > >> index 000..f6e3e37 > >> ---

Re: [libvirt] [PATCH v2] Add support for Veritas HyperScale (VxHS) block device protocol

On Sat, Jan 14, 2017 at 10:31:05AM -0500, John Ferlan wrote: > [...] > > >> diff --git > >> a/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-vxhs.args > >> b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-vxhs.args > >> new file mode 100644 > >> index 000..f6e3e37 > >> ---

Re: [libvirt] How to generate better API documentation?

On 16.01.2017 08:24, Michal Privoznik wrote: > Dear list, > > now that we more or less agreed to use some new features (i.e. automatic > free() when a variable goes out of the scope [1]), and with our NEWS efforts, > do you think it is finally the right time to fix generation of our API docs >