SUMMARY:Automounting C Compiler/Sol.5

From: Lau, Victoria H (vlau@msmail2.hac.com)
Date: Tue Jul 30 1996 - 23:24:28 CDT


Hello everybody:

Since I posted the following summary on automounting c compiler, I have
received more info from Kenneth Simpson and Michael J. Shon which I would
like to forward to everyone. By the way, thank you both for taking the time
to enlighten me.

Hi Vicky - incidently, the warning is mainly for diskless clients. A hard
mount
is a mount with a timeout=infinity and a soft mount a mount with a timeout=10
seconds.
You can vary the the timeout period of a soft mount to be anything between 0
and
infinity. In addition, if you're reading the binaries into memory and an I/O
occurs, it will indicate there was an error. No big deal. Or if you can't
write
an object file and I/O occurs - it will complain - no big deal. The problem
is
with applications which don't indicate there's an I/O error - usually home
baked. In short, if you have trouble with soft mounts then you probably have
other network problems too. In short, it's not a 1 or 0 issue - or a black and

white issue - it's a grey issue and you need to experiment to figure out what
works for you.

-- Ken

|server is up. I did not use the "soft" mount because the man page on
|mount_nfs warns that "File systems that are mounted read-write or that
|contain executable files should always be mounted with the hard option.
|Applications using soft mounted file systems may incur unexpected I/O
|errors."
|

For long-running apps (like window managers), you do not want soft mounts,
but for a thing like a compiler, it should be fine.
If the server goes down while you are compiling, you probably wouldn't mind
having your "make" fail as long as you could get your prompt back.

    Michael Shon (716) 385-5065 michael.shon@East.Sun.COM

_______________________________________________________________________________
From: Lau, Victoria H on Wed, Jul 24, 1996 8:32 PM
Subject: SUMMARY:Automounting C Compiler/Sol.5
To: sun-managers@ra.mcs.anl.gov

Hi everybody:

I'd posted a question for help on automounting C Compiler on systems running
Solaris 2.5. When users are accessing this automounted file system and
the server goes down, the session hangs until the server comes back up. I
was looking for ways to ^C out of th session so that the users can "move on
with their lives," especially for those users on the VT100 terminals.

I'd received one response from Kenneth Simpson (thank you!) who suggested
that I should "soft" mount the c compiler directory and avoid settig the
search paths to this automounted file system in .cshrc.

I took the advise and created a wrapper for each of the executables (like
cc, CC, visu, small_visu, etc.) which sets up the paths AFTER testing the
server is up. I did not use the "soft" mount because the man page on
mount_nfs warns that "File systems that are mounted read-write or that
contain executable files should always be mounted with the hard option.
Applications using soft mounted file systems may incur unexpected I/O
errors."

I'm using the following line in my /etc/auto_direct file to mount
the C Compiler file system from the server:

  /opt/vws_v2.1 -ro,intr,retry=0 r2cp2:/opt/vws_v2.1

However, Sun told me that this line is read only during the mount, and
once the file system has been mounted/automounted, there is no way to
set a time-out period nor interrupt it with ^C. The only workaround is
logging in from another terminal/window and use "ps" to kill the process.

I'm sorry that I can't include my original question nor Ken's response
because my PC crashed two weeks ago and all the emails I downloaded to my
hard drive are gone...and so are all the SUMMARIES I collected over the
year. I know I promised a few people to send them the logs on automating
diskless client install and customizing Solaris 2.5 cd using Young Minds.
I may post these to the group as SUMMARY in a couple of weeks since I lost
all of the names and addresses.

Sincerely,
Vicky Lau
vlau@msmail2.hac.com

------------------ RFC822 Header Follows ------------------
Received: by MSMAIL2.HAC.COM with SMTP;24 Jul 1996 20:32:16 -0800
Received: from 140.221.4.26 by EDEN.HAC.COM (PMDF V4.3-7 #5884)
 id <01I7GPTAUW3K00AAU4@EDEN.HAC.COM>; Wed, 24 Jul 1996 20:30:03 PST
Received: from localhost (daemon@localhost) by ra.mcs.anl.gov (8.7.4/8.7.3)
 with SMTP id WAA13602; Wed, 24 Jul 1996 22:04:11 -0500 (CDT)
Received: by ra.mcs.anl.gov (bulk_mailer v1.5); Wed, 24 Jul 1996 22:01:01
-0500
Received: (from daemon@localhost) by ra.mcs.anl.gov (8.7.4/8.7.3)
 id VAA13585 for sun-managers-outbound; Wed, 24 Jul 1996 21:58:06 -0500 (CDT)
Received: from igate1.hac.com (igate1.HAC.COM [192.48.33.10])
 by ra.mcs.anl.gov (8.7.4/8.7.3) with SMTP id VAA13580 for
 <sun-managers@ra.mcs.anl.gov>; Wed, 24 Jul 1996 21:58:03 -0500 (CDT)
Received: from msmail2.hac.com ([147.17.106.40])
 by igate1.hac.com (4.1/SMI-4.1) id AA14392; Wed, 24 Jul 96 19:58:32 PDT
Date: Wed, 24 Jul 1996 20:00:43 -0800
From: "Lau, Victoria H" <vlau@msmail2.hac.com>
Subject: SUMMARY:Automounting C Compiler/Sol.5
Sender: sun-managers-relay@ra.mcs.anl.gov
To: sun-managers@ra.mcs.anl.gov
Reply-to: "Lau, Victoria H" <vlau@msmail2.hac.com>
Message-id: <n1373883901.12217@MSMAIL2.HAC.COM>
X-Mailer: Mail*Link SMTP-MS 3.0.2
Content-transfer-encoding: 7BIT
Precedence: bulk
Followup-to: junk



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:11:06 CDT