[Qemu-devel] Re: [kvm-devel] Feedback and errors

[Avi Kivity]

Anthony Liguori wrote:
> > > 2/ two instances of kvm can be passed the same -hda. There is no 
> > > locking whatsoever. This messes up things seriously.
> > >
> > >       
> >
> > These two are upstream qemu problems. Copying qemu-devel.
> >
> > I guess using file locking by default would improve the situation, 
> > and we can add a -drive ...,exclusive=no option for people playing 
> > with cluster filesystems.
> >   
>
> This is not a situation where the user has a reasonable expectation of 
> what will happen that we violate.  If the user is unhappy with the 
> results, it's because the user made a mistake.  

Well, one user (me) has made this mistake, several times.

> FWIW, the whole override thing for Xen has been an endless source of 
> pain.  It's very difficult (if not impossible) to accurately determine 
> if someone else is using the disk.  

What's wrong with the standard file locking API?  Of course it won't 
stop non-qemu apps from accessing it, but that's unlikely anyway.

> Also, it tends to confuse people trying to do something legitimate 
> more often than helping someone doing something stupid.

-drive exclusive=off (or share=yes)

> I very frequently run multiple VMs with the same disk.  I do it 
> strictly for the purposes of benchmarking.  There are ways to share a 
> disk without using a clustered filesystem.

I imagine only raw format disks, and only as non-root filesystems (or 
with -shapshot, which should automatically set exclusive=off)?

> If a higher level management tool wants to enforce a policy (like 
> libvirt), then let it.  We should not be enforcing policies within 
> QEMU though.

I agree that qemu is not the place to enforce policies, but covering a 
hole that users are likely to step into, while allowing its explicit 
uncovering, is a good thing.  We're not enforcing the policy, only hinting.


--
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.