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

Re: [Xen-devel] Migration memory corruption - PV backends need to quiesce

At 12:01 +0100 on 30 Jun (1404126084), Ian Campbell wrote:
> On Mon, 2014-06-30 at 12:14 +0200, Tim Deegan wrote:
> > Yes.  The (suggested) problem is that live migration does not preserve
> > that write ordering.  So we have to worry about something like this:
> > 
> > 1. Toolstack pauses the domain for the final pass.  Reads the final
> >    LGD bitmap, which happens to include the shared ring but not the
> >    data pages.
> WRT this -- writes by the backend don't cause the page to become dirty.
> What about writes by the f.e? Are pages which are granted and actively
> mapped by another domain handled as normal guest RAM pages for the
> purposes of logdirty or are they handle specially?

They're handled as normal RAM, so yes f/e writes will mark that page


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.