The design document for mirror-mounts says: For NFSv4-capable servers (only), providing the client's "/net" functionality by the mirror-mount mechanism, instead of by the automounter.
Does this mean that mm won't provide the inital /net/hostname mount but will still solve the problem of multiple nested filesystems that appear over time eg ? /net/hostname/zbuilds/username/build1 /net/hostname/zbuilds/username/build2 Where zbuilds zbuilds/username zbuilds/username/build? are all separate file systems. If I then later on the server create /net/hostname/zbuilds/username/build3 Then the mm functionality on the NFS client will give me build3. Am I understanding this correctly ? -- Darren J Moffat This message posted from opensolaris.org
