Re: nfs patch for NetApp snapshot

James H. Cloos Jr. (cloos@jhcloos.com)
19 Dec 1998 06:31:07 -0600


>>>>> "Chris" == Chris Wedgwood <cw@ix.net.nz> writes:

Chris> Why do we have a specail hardcoded case for '.snapshot' files?
Chris> I probably don't know enough about how NetApps work, but this
Chris> seems rather unusual... why is this magic required to make this
Chris> work?

Becase (where $MNT is the mountpoint):

$MNT/foo/bar
$MNT/.snapshot/$SNAPNAME/foo/bar
$MNT/foo/.snapshot/$SNAPNAME/bar

all have the same inode number, but may have different values for the
inode structures and/or user data.

One problem I've seen with earlier kernels due to this, for instance,
is that accessing a snapshot can have the effect that is the logical
equivalent of remounting the snapshot over the main remote directory;
suddenly the filesystem is rolled back and, if originally mounted rw,
is suddenly ro. The only fix is to invalidate the cache. On a busy
server that generally equated to a reboot.

-JimC

-- 
James H. Cloos, Jr.       <http://www.jhcloos.com/cloos/pgp_public_key.txt>
<cloos@jhcloos.com>      E9E9 F828 61A4 6EA9 0F2B  63E7 997A 9F17 ED7D AEA6

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/