Re: [PATCH RFC] memory: Don't allow to resize RAM while migrating

["Dr. David Alan Gilbert" <dgilbert@redhat.com>]

* David Hildenbrand (david@redhat.com) wrote:
> On 14.02.20 11:25, Dr. David Alan Gilbert wrote:
> > * David Hildenbrand (david@redhat.com) wrote:
> >> Resizing while migrating is dangerous and does not work as expected.
> >> The whole migration code works on the usable_length of ram blocks and does
> >> not expect this to change at random points in time.
> >>
> >> Precopy: The ram block size must not change on the source, after
> >> ram_save_setup(), so as long as the guest is still running on the source.
> >>
> >> Postcopy: The ram block size must not change on the target, after
> >> synchronizing the RAM block list (ram_load_precopy()).
> >>
> >> AFAIKS, resizing can be trigger *after* (but not during) a reset in
> >> ACPI code by the guest
> >> - hw/arm/virt-acpi-build.c:acpi_ram_update()
> >> - hw/i386/acpi-build.c:acpi_ram_update()
> >>
> >> I see no easy way to work around this. Fail hard instead of failing
> >> somewhere in migration code due to strange other reasons. AFAIKs, the
> >> rebuilts will be triggered during reboot, so this should not affect
> >> running guests, but only guests that reboot at a very bad time and
> >> actually require size changes.
> >>
> >> Let's further limit the impact by checking if an actual resize of the
> >> RAM (in number of pages) is required.
> >>
> >> Don't perform the checks in qemu_ram_resize(), as that's called during
> >> migration when syncing the used_length. Update documentation.
> > 
> > Interesting; we need to do something about this - but banning resets
> > during migration is a bit harsh; and aborting the source VM is really
> > nasty - for a precopy especially we shouldn't kill the source VM,
> > we should just abort the migration.
> 
> Any alternative, easy solutions to handle this? I do wonder how often
> this will actually trigger in real life.

Well it's not that hard to abort a migration (I'm not sure we've got a
convenient wrapper to do it - but it shouldn't be hard to add).

> > 
> > The other thing that worries me is that acpi_build_update calls
> >    acpi_ram_update->memory_region_ram_resize
> > multiple times.
> 
> It's different memory regions, no? table_mr, rsdp_mr, linker_mr.

Oh, so it is.

> > So, it might be that the size you end up with at the end of
> > acpi_build_update is actually the same size as the original - so
> > the net effect is the RAMBlock didn't really get resized.
> 
> Are you sure?

No!
Avocado has a migration+reset test, so it's worth trying.
Certainly in a cloud setup migrations happen often and no one knows
what the guest is doing;  aborting the source isn't acceptable.

It surprises me a bit that the region sizes would change due to guest
actions - I thought they were determined by the set of virtual hardware;
not sure if a hot-unplug/plug followed by reset would trigger it or not.

Dave

> 
> -- 
> Thanks,
> 
> David / dhildenb
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK