Hello,
one of NetBSD's Summer-of-Code projects which wasn't taken on by
student so far is the "Improved Automounter Support" project:
NetBSD currently uses The Berkeley Automounter Suite of Utilities for
p automatically mounting (network) file systems. This software package
implements an automounter file system as a userland NFS daemon. While
this generally works it has major drawbacks:
* File systems are not mounted directly on the desired mount point.
As a result applications frequently use incorrect pathnames (e.g.
"/amd/server/home/user" instead of "/home/user") for automatically
mounted directories or files beneath them. This is especially problematic
in heterogeneous enviroments where not all machines use the same automounter.
* The automounter daemon cannot handle high I/O load very well, file access
occasionally fails with intermittent errors. Please look at this Pr
for more details: http://gnats.netbsd.org/41259
I can think of two ways to implement a better automounter daemon:
1.) Implement a kernel automounter file-system similar to what
Solaris and Linux are using. While this should definitely
work (as proven by the other implementations) I will however
not be a very easy task.
2.) Use Antti Kantee's most excellent "Pass-to-Userspace Framework
File System" (see http://www.netbsd.org/docs/puffs/) to
implement a new automounter as a userspace application.
This should be much better managable as one doesn't have
to deal with kernel restrictions (e.g. limitted stack space),
allow for easier debugging and avoid development delays
caused by reboots (panics, new kernels, etc.).
I would greatly appreciate a student which would like to work on this
project during the Google Summer-of-Code 2010. NetBSD's current
automounter is something which (not only) I have been unhappy
about for quite a while.
Kind regards
--
Matthias Scheler http://zhadum.org.uk/
Attachment:
pgp6VRzhUNY9I.pgp
Description: PGP signature