Automount question

[gaf-mNDKBlG2WHs at public.gmane.org (Jerry Feldman)]

On 08/09/2010 05:31 PM, Edward Ned Harvey wrote:
>> From: discuss-bounces-mNDKBlG2WHs at public.gmane.org [mailto:discuss-bounces-mNDKBlG2WHs at public.gmane.org] On
>> Behalf Of Jerry Feldman
>>
>> In my auto.master I have=20
>> /mnts   /etc/auto.mnts
>> My /etc/auto.mnts is
>> foo    -fstype=3Dnfs,rw,nosuid  <host0>:/exports/foo
>>
>> This works fine, but for a transition period I will be moving some of
>> foo's subdirectories individually so I want to automount
>> <host1>:/mnts/foo/clients
>>    =20
> I would recommend using direct automount.  It's a new feature in autofs=
5,
> but autofs5 is the default in anything more recent than RHEL4.  Even if=
 you
> have RHEL4, you can install the updated autofs5 package, and I always d=
o.
> Specifically because direct automount is do *darn* much better than the=
 old
> way, you're describing.
>
> Here's an example direct automount config:
>
> # Contents of /etc/auto.master
> /-  /etc/auto.direct --timeout=3D1200
>
> # Contents of /etc/auto.direct
> /any/mount/path  -fstype=3Dnfs,noacl,rw,hard,intr,posix
> fileserver:/some/export/path
>  =20
The bottom line is that I set up auto.direct and it works like a charm
on all servers. We also got another benefit. A year or so ago (Jan
2009), we had a discussion on memmap locking where one of our
applications reported an mmap locking failure, but it was the only
process locking the file. That was reported as a kernel bug, but we
can't change our running kernel. I changed some of the mount options,
and we no longer have that problem. In any case direct automount works.
What is interesting is we are using RHEL 5.2, and the kernel reports
"starting aufofs4'.

--=20
Jerry Feldman <gaf-mNDKBlG2WHs at public.gmane.org>
Boston Linux and Unix
PGP key id: 537C5846
PGP Key fingerprint: 3D1B 8377 A3C0 A5F2 ECBB  CA3B 4607 4319 537C 5846