Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 7 Jun 1996 15:50:41 -0600
From:      Nate Williams <nate@sri.MT.net>
To:        "Karl Denninger, MCSNet" <karl@mcs.com>
Cc:        pst@shockwave.com (Paul Traina), security@FreeBSD.org
Subject:   Re: FreeBSD's /var/mail permissions
Message-ID:  <199606072150.PAA00803@rocky.sri.MT.net>
In-Reply-To: <m0uS9Nx-000IDOC@venus.mcs.com>
References:  <199606072105.OAA00533@precipice.shockwave.com> <m0uS9Nx-000IDOC@venus.mcs.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Karl Denninger writes:
> > Mail locking, to be effective, must be soley performed through the use of
> > the flock() call on the mail file itself.
> > 
> > Locking schemes relying on other mechanisms are not effective.
> > 
> > Sorry.
> 
> Mail locking, to be effective, must *work across machines* so that NFS
> mounts of the mail directory work.
> 
> flock() cannot be trusted to work in this environment, and in fact doesn't
> even attempt to work on FreeBSD.
> 
> Does this mean we should give up on using mail?

Actually, *nothing* works reliably across NFS locks if you don't have
lockd, and since FreeBSD doesn't (yet) have lockd you shouldn't have
mailboxes that can be written to by multiple processes on different
systems.

There are *VERY* few systems (including most of Sun's) that implement
reliable NFS file-locking well.  Sun attempted it many times and
failed. :(




Nate



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199606072150.PAA00803>