similar to: lvm vs qcow (on NVMe)

On Fri, Oct 26, 2018 at 12:32:48PM +0100, lejeczek wrote: > hi guys, > > anybody observed, maybe has measured performance difference between > the two with qemu quests(windows OS)? > > Any conclusions, recommendations? [Cc Stefan from QEMU, probably he might have some pointers.] -- /kashyap

Hi All I was reading through this: http://kashyapc.com/2011/12/02/little-more-disk-io-perf-improvement-with-fallocateing-a-qcow2-disk/ I was basically searching for pointers on improving disk I/O. I wanted to know the purpose of preallocation=full & preallocation=metadata , What is the difference between them ? and which one would yield a better disk I/O speed ? Appreciate your

Hello, I am running several virtualization servers with QEMU 1.4.x and libvirt 1.0.2 on Ubuntu 12.04 and am working on optimizing the cache= and aio= options for the virtual machines. These VM images are mostly qcow2, and are served both from a local ext4 filesystem (with data=ordered,barrier) and from an NFS mountpoint (with sync). The local filesystem sits on top of an md software RAID of SATA

Hi, Thanks Kashyap for your answer. I am trying to give more data: virt-install invokation: # virt-install --name eff1 --disk /home/kvmimages/eff1.img --cdrom /home/kvmimages/Fedora-20-x86_64-DVD.iso --memory 4096 WARNING CDROM media does not print to the text console by default, so you likely will not see text install output. You might want to use --location. Début d'installation...

hi guys, I thought I should ask before I start fiddling and break something - can we tell grub2 to add nvme device as an option to boot from? And if yes then what do tell grub? many thanks, L.

On Mon, Dec 17, 2018 at 01:52:40PM -0500, Jonathan Billings wrote: > > On Mon, Dec 17, 2018 at 05:50:44PM +0000, lejeczek via CentOS wrote: > > I updated to qemu-kvm-common-ev-2.12.0-18.el7_6.1.1.x86_64 (also libvirt) > > and my Centos 6.10 kvm guests now do not start. > > > > Funnily enough Win10 guest are fine, only Centoses cannot start, silently & > >

Am 10.06.13 09:07, schrieb Kashyap Chamarthy: > On 06/09/2013 02:46 PM, Thomas Stein wrote: >> Hello. >> >> I just tried the following: >> >> virsh dumpxml --security-info gentoo-template > gentoo-template.xml >> virsh snapshot-create gentoo-template >> virsh undefine gentoo-template > > Wait - is that really the command you used after

Thanks for your interest. Here are the dir/files permissions for the kvmimages dir: # ls -ld /home /home/kvmimages /home/kvmimages/* drwxr-xr-x. 4 root root 4096 14 juin 23:05 /home drwxrwxrwx 2 root root 4096 15 juin 09:19 /home/kvmimages -rw-rw-rwx 1 root root 12884901888 14 juin 23:07 /home/kvmimages/eff1.img -rw-rw-rwx 1 root root 12884901888 14 juin 23:11

Hi I have reinstall Fedora 20 from distrib. I have run yum update and yum upgrade. Running kernel is : 3.14.7-200.fc20.x86_64 Installed Libvirt packages are: libvirt-client-1.1.3.5-2.fc20.x86_64 libvirt-daemon-1.1.3.5-2.fc20.x86_64 libvirt-daemon-config-network-1.1.3.5-2.fc20.x86_64 libvirt-daemon-driver-interface-1.1.3.5-2.fc20.x86_64 libvirt-daemon-driver-network-1.1.3.5-2.fc20.x86_64

On Tue, Dec 09, 2014 at 05:48:04PM +0100, Kashyap Chamarthy wrote: > On Tue, Dec 09, 2014 at 11:27:02AM +0000, Richard W.M. Jones wrote: > > On Tue, Dec 09, 2014 at 10:53:41AM +0000, Peter Maydell wrote: > > > On 9 December 2014 at 10:50, Kashyap Chamarthy <kchamart@redhat.com> wrote: > > > > Booting a minimal KVM guest throws the below error on Cubietruck: >

Running virt-builder in a guest Fedora-20 guest hypervisor w/ Rawhide Kernel, throws the below: $ virt-builder fedora-20 --format qcow2 --size 20G virt-builder: symbol lookup error: /lib64/libgnutls.so.28: undefined symbol: nettle_secp_256r1 Version on L1 (guest hypervisor) -------------------------------- $ uname -r; rpm -q libvirt-daemon-kvm qemu-system-x86 \ libguestfs gnutls

We're currently support two paths from VFS to preallocate unwritten extents(from FS_IOC_RESVSP, or fallocate()), likewise, behavior of punching-hole should be treated as the same, this patch tries to teach file_ioctl() to handle FS_IOC_UNRESVSP, underlying filesystem like ocfs2 is wise enough to do the rest of work;-) Signed-off-by: Tristan Ye <tristan.ye at oracle.com> --- fs/ioctl.c

We're currently support two paths from VFS to preallocate unwritten extents(from FS_IOC_RESVSP, or fallocate()), likewise, behavior of punching-hole should be treated as the same, this patch tries to teach file_ioctl() to handle FS_IOC_UNRESVSP, underlying filesystem like ocfs2 is wise enough to do the rest of work;-) Signed-off-by: Tristan Ye <tristan.ye at oracle.com> --- fs/ioctl.c

Cc the QEMU Block Layer mailing list (qemu-block@nongnu.org), who might have more insights here; and wrap long lines. On Mon, May 28, 2018 at 06:07:51PM +0800, Chunguang Li wrote: > Hi, everyone. > > Recently I am doing some tests on the VM storage+memory migration with > KVM/QEMU/libvirt. I use the following migrate command through virsh: > "virsh migrate --live

On Mon, May 25, 2015 at 10:22:35PM +0200, Kashyap Chamarthy wrote: > I just built `supermin` from source on my Fedora 22 machine, and somehow > it can't seem to detect RPM (but `supermin` from Rawhide detects it). > Exact error: > > supermin: could not detect package manager used by this system or > distro. > > which is coming from:

I'm on libguestfs Fedora git master branch: $ git log | head -1 commit 8eb1aa2406632eb4202bbd976459334854295a77 $ yum-builddep libguestfs $ fedpkg local Results in: ========== . . . make[4]: Leaving directory `/home/kashyap/fedora-packaging/libguestfs/master/libguestfs-1.25.18/gobject' Making all in docs make[4]: Entering directory

On Fri, Apr 03, 2015 at 10:13:38AM +0200, Kashyap Chamarthy wrote: > On Fri, Apr 03, 2015 at 10:08:21AM +0200, Kashyap Chamarthy wrote: > > Migration without --p2p works just fine, ie. the below works: > > > > > > $ virsh migrate --verbose --copy-storage-all \ > > --live cvm1 qemu+ssh://kashyapc@devstack3/system > > Migration: [100 %] >

Il 13/01/2015 10:51, Kashyap Chamarthy ha scritto: > On Tue, Jan 13, 2015 at 10:10:53AM +0100, Fiorenza Meini wrote: >> Hi there, >> I receive this error when I run nova image-create <VM name> <Vm Sanpshot >> name>: > > Okay, you're talking in the context of OpenStack. > > You can also check the Nova compute.log for more contextual details of >

On 07.02.2018 16:31, Kashyap Chamarthy wrote: > [Cc: KVM upstream list.] > > On Tue, Feb 06, 2018 at 04:11:46PM +0100, Florian Haas wrote: >> Hi everyone, >> >> I hope this is the correct list to discuss this issue; please feel >> free to redirect me otherwise. >> >> I have a nested virtualization setup that looks as follows: >> >> - Host:

On 01/17/2014 04:06 PM, Rom Freiman wrote: > Kashyap, just to be sure - it happens to you during the v2v > conversion? on L2? I haven't done any v2v conversions in L2 (or at any other level). PS: Sorry, I didn't notice my previous 2 emails didn't go to the list, that wasn't intended. Rich, you bounce them here, if you prefer (instead of me clumsily forwarding them). --