Am-utils Frequently Asked Questions

Note: we started this FAQ only on March 15, 2005; so it's not long or
comprehensive, yet.  Amd is much older than that, and so there's a lot of
information that's already available in other forms.  If this FAQ doesn't
answer your questions, see information in the following sources:

1. The Am-utils book: http://www.am-utils.org/docs/amd-book/

2. The Am-utils user manual, which is part of the distribution and is also
   available from www.am-utils.org.

3. The www.am-utils.org Web site resources, especially the "am-utils" mailing
   list (and its archives).

4. In the am-utils distribution (always use the latest ones), see all of the
   various README files (README, README.autofs, README.ldap, README.osx, and
   README.y2k).  The "BUGS" file also lists useful information about bugs
   and problems with specific OSs which affect Amd.  All of these text files
   are also available from www.am-utils.org.

5. Some FAQ questions (including newbie questions) are available here:
   http://www.kernelcorp.com/resources_faqs.html

6. Some problems are known bugs but have not been fixed yet: this are
   listed in bugzilla in https://bugzilla.am-utils.org/

If you have additions to this FAQ, please let us know at
the am-utils list (see www.am-utils.org).

Thank you,
The Am-utils development team.

<FAQ>

*** Linux Questions

Q1. When I use Amd with Autofs and I restart Amd, how come it cannot remount
    the Autofs partitions?

A1. This is a limitation of the Linux Autofs kernel module (for both autofs
    v2. and v3).  The Linux Autofs does not allow restarting automounted
    points.  There's nothing Amd can do about this.  In fact, the same
    problem exists if you use the userland "automount" daemon instead of
    Amd.  Hopefully Autofs-v4 or the separate effort of Autofs-NG will
    address this serious problem.

    Note that Amd itself can restart autofs automounted points just fine on
    OSs that support it, for example Solaris.


Q2. When I use Amd, I get this console message frequently: "mount version
    older than kernel."  Is it a problem?

A2. No, it's a harmless warning message that the Linux kernel prints for NFS
    mounts.  The intent was to alert administrators that the kernel has
    supposedly a different version of the mount(2) code than a userland
    program used.  This happens if you compile Amd against kernel headers
    that are different than the kernel you're running.  If the message
    really bothers you, then one way to "fix" the problem is to recompile
    Amd against the same kernel headers as the running kernel.

    Nevertheless, it is a relatively useless message because as far as we
    know, the NFS v2 and v3 mount codes have been in perfect sync between
    the userland and kernel sides, and were "standardized" for years
    already.  This warning message caused more unnecessary worry among
    administrators than helping alert them to legitimate problems.

</FAQ>