qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH 0/3] Introduce a new --only-migratable option


From: Dr. David Alan Gilbert
Subject: Re: [Qemu-devel] [PATCH 0/3] Introduce a new --only-migratable option
Date: Thu, 15 Dec 2016 19:03:56 +0000
User-agent: Mutt/1.7.1 (2016-10-04)

* Michael S. Tsirkin (address@hidden) wrote:
> On Thu, Dec 15, 2016 at 04:07:41PM +0000, Dr. David Alan Gilbert wrote:
> > * Michael S. Tsirkin (address@hidden) wrote:
> > > On Thu, Dec 15, 2016 at 12:36:59AM +0530, Ashijeet Acharya wrote:
> > > > This series adds a new command line option "--only-migratable" which 
> > > > will only
> > > > allow addition of those devices to a QEMU instance which are migratable 
> > > > and do
> > > > not abruptly fail QEMU after migration.
> > > > 
> > > > Patch 1 adds the new option "-only-migratable".
> > > > 
> > > > Patch 2 adds compatibility for various "device adding" options for both 
> > > > via
> > > > command line and hotplug methods.
> > > > 
> > > > Patch 3 handles the special case of devices which become unmigratable 
> > > > dynamically
> > > > by making call to "migrate_add_blocker". Here we fail the particular 
> > > > action of the
> > > > device which results in an unmigratable VM.
> > > > Eg: 9pfs fails to mount the filesystem.
> > > 
> > > I guess that's possible but wouldn't it be cleaner to add introspection
> > > so management can find out what can migrate?
> > > Further, what should be removed (e.g. by hotplug) if you do want
> > > to migrate.
> > > 
> > > As it is, you get a failure but no easy way for tools to
> > > find out what was the failure reason.
> > 
> > You get the device name of the device that's failed don't you?
> 
> As a string but I'm not sure it's machine readable.
> 
> > Where would you add introspection?
> > 
> > Dave
> 
> monitor and command line.

OK; that's not a bad idea if you or a libvirt person could
specifiy how they want that to appear - but it's really
a separate problem.
One thing we'd have to do if we did that would be to
add another flag to devices to declare that they might
sometime in the future call the migrate_add_blocker.

Dave

> > > 
> > > > Ashijeet Acharya (3):
> > > >   migration: Add a new option to enable only-migratable
> > > >   migration: Allow "device add" options to only add migratable devices
> > > >   migration: disallow migrate_add_blocker during migration
> > > > 
> > > >  block/qcow.c                  | 11 ++++++++++-
> > > >  block/vdi.c                   | 11 ++++++++++-
> > > >  block/vhdx.c                  | 20 ++++++++++++++------
> > > >  block/vmdk.c                  | 12 +++++++++++-
> > > >  block/vpc.c                   | 15 ++++++++++++---
> > > >  block/vvfat.c                 | 24 ++++++++++++++++--------
> > > >  hw/9pfs/9p.c                  | 22 ++++++++++++++++++----
> > > >  hw/display/virtio-gpu.c       | 35 ++++++++++++++++++++++-------------
> > > >  hw/intc/arm_gic_kvm.c         | 20 ++++++++++++++------
> > > >  hw/intc/arm_gicv3_its_kvm.c   | 21 ++++++++++++++-------
> > > >  hw/intc/arm_gicv3_kvm.c       | 22 +++++++++++++++-------
> > > >  hw/misc/ivshmem.c             | 17 +++++++++++++----
> > > >  hw/scsi/vhost-scsi.c          | 27 +++++++++++++++++++++------
> > > >  hw/usb/bus.c                  | 15 +++++++++++++++
> > > >  hw/virtio/vhost.c             | 11 ++++++++++-
> > > >  include/migration/migration.h |  9 ++++++++-
> > > >  migration/migration.c         | 42 
> > > > ++++++++++++++++++++++++++++++++++++++++--
> > > >  qdev-monitor.c                |  9 +++++++++
> > > >  qemu-options.hx               | 10 ++++++++++
> > > >  stubs/migr-blocker.c          |  3 ++-
> > > >  target-i386/kvm.c             | 19 ++++++++++++++++---
> > > >  vl.c                          |  4 ++++
> > > >  22 files changed, 304 insertions(+), 75 deletions(-)
> > > > 
> > > > -- 
> > > > 2.6.2
> > --
> > Dr. David Alan Gilbert / address@hidden / Manchester, UK
--
Dr. David Alan Gilbert / address@hidden / Manchester, UK



reply via email to

[Prev in Thread] Current Thread [Next in Thread]