help-hurd
[Top][All Lists]
Advanced

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

Re: Running several instances of an fs server...


From: Marcus Brinkmann
Subject: Re: Running several instances of an fs server...
Date: Wed, 24 Jul 2002 03:00:01 +0200
User-agent: Mutt/1.4i

On Tue, Jul 23, 2002 at 10:25:13AM -0700, Thomas Bushnell, BSG wrote:
> Marcus Brinkmann <Marcus.Brinkmann@ruhr-uni-bochum.de> writes:
> 
> > I should add that of course it is technically possible to implement a much
> > more complex protocol that allows a filesystem that starts up to find out
> > that the device (really: storage) is currently used by another filesystem of
> > the same type and ask that to "take over".  However, such a solution is not
> > part of the current design, and a bit too specific to justify the burden of
> > its implementation.  Compare it with editors and shared text files.  If you
> > need a custom solution to allow concurrent editing of the same file, you
> > write one (like, a special emacs mode), you don't make it a feature of the
> > operating system.
> 
> Um, Marcus, this "much more complex protocol" is just flock.

We could do better than flock.  flock can be used to just protect against the
simultaneous use of the same storage for two filesystems, but it could only
be used to detect the situation and throw an error if it occurs.  However, I
spoke about "take over" above on purpose.
 
What I had in mind is something more advanced.  Consider two users
want to use ext2fs on /dev/hd0s1.  Then the more advanced protocol would let
the same translator service requests on both "mount points".

Imagine the following sequence of commands:

settrans -a /mnt /hurd/ext2fs /dev/hd0s1
settrans -a /mnt2 /hurd/ext2fs /dev/hd0s1

While the first one would install a real translator, the second one would,
in the more advanced scheme, notice that another instance of ext2fs with the
same options and backing store is already running and would just ask the
running translator to also attach itself to /mnt2.  At least this is what I
meant, I don't really think it makes sense to do it this way.  In fact, I
think there are many reasons not to do it this way.

Thanks,
Marcus

-- 
`Rhubarb is no Egyptian god.' GNU      http://www.gnu.org    marcus@gnu.org
Marcus Brinkmann              The Hurd http://www.gnu.org/software/hurd/
Marcus.Brinkmann@ruhr-uni-bochum.de
http://www.marcus-brinkmann.de/



reply via email to

[Prev in Thread] Current Thread [Next in Thread]