]> xenbits.xensource.com Git - people/liuw/libxenctrl-split/libvirt.git/commit
qemu: Memory pre-pinning support for RDMA migration
authorMichael R. Hines <mrhines@us.ibm.com>
Mon, 13 Jan 2014 06:28:12 +0000 (14:28 +0800)
committerJiri Denemark <jdenemar@redhat.com>
Tue, 23 Sep 2014 06:11:50 +0000 (08:11 +0200)
commit9cc1586d2b5b6d4d396ebfa0479ae441e0b917ea
treec09bec7bfb3665a160f77c25447ee6431984039e
parented22a4743437b6f646eb33659a212dd252a5b65d
qemu: Memory pre-pinning support for RDMA migration

RDMA Live migration requires registering memory with the hardware, and
thus QEMU offers a new 'capability' to pre-register / mlock() the guest
memory in advance for higher RDMA performance before the migration
begins. This capability is disabled by default, which means QEMU will
register the memory with the hardware in an on-demand basis.

This patch exposes this capability with the following example usage:

virsh migrate --live --rdma-pin-all --migrateuri rdma://hostname domain qemu+ssh://hostname/system

Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
include/libvirt/libvirt.h.in
src/qemu/qemu_migration.c
src/qemu/qemu_migration.h
tools/virsh-domain.c