On Thu, Sep 07, 2006 at 01:52:33PM -0700, Darren Dunham wrote: > > What are the problems that you see with that check? It appears similar > to what VxVM has been using (although they do not use the `hostid` as > the field), and that appears to have worked well in most cases. > > I don't know what issues appear with multiple hosts. My worry is that > an accidental import would allow two machines to update uberblock and > other metadata to the point that you get corruption. If the "sharing" > hosts get read-only access and never touch the metadata, then (for me) > the hostname check becomes much less relevant. If they want to import > it, fine...but don't corrupt anything. >
I agree that it's a useful check against accidental mistakes - as long we're not talking about some built in clustering behavior. We just haven't thought through what the experience should be. In particular, there are some larger issues in relation to FMA that need to be address. For example, we would want the pool to show up as faulted, but there needs to be a consistent way to 'repair' such a pool. Should it be an extension of 'zpool clear', or should it be done through 'fmadam repair'? Or even 'zpool export' followed by 'zpool import'? We're going to have to answer these questions for the next phase of ZFS/FMA interaction, so maybe it would be a good time to think about this problem as well. And of course, you'll always be able to shoot yourself in the foot if you try, either by 'repairing' a pool that's actively shared, or by force-importing a pool that's actively in use somewhere else. - Eric -- Eric Schrock, Solaris Kernel Development http://blogs.sun.com/eschrock _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
