From:	MERC::"uunet!ARISIA.dnet.ge.com!CRDGW2::CRDGW2::MRGATE::SMTP::CRVAX.SRI.COM::RELAY-INFO-VAX" 13-APR-1992 14:48:16.90
To:	galaxy::GleEve
CC:	
Subj:	Re: TGV MultiNet vs. WIN/TCP NFS Client

From:	RELAY-INFO-VAX@CRVAX.SRI.COM@SMTP@CRDGW2
To:	Everhart@Arisia@MRGATE

Received:  by crdgw1.ge.com (5.57/GE 1.137)
	 id AA00147; Mon, 13 Apr 92 07:47:17 EDT
Received: From TGV.COM by CRVAX.SRI.COM with TCP; Mon, 13 APR 92 00:57:30 PDT
Date: Mon, 13 Apr 1992 00:57:05 -0700 (PDT)
From: KASHTAN@TGV.COM (David L. Kashtan)
Subject: Re: TGV MultiNet vs. WIN/TCP NFS Client
To: smith@mcclb0.med.nyu.edu
Cc: wintcp-l%ubvm.bitnet@cunyvm.cuny.edu, info-vax@sri.com,
        info-multinet@TGV.COM
Message-Id: <703151825.374000.KASHTAN@TGV.COM>
In-Reply-To:  <1992Apr11.172442.7081@mcclb0.med.nyu.edu>
Mail-System-Version: <VAX-MM(309)+TOPSLIB(155)@TGV.COM>
Postal-Address: TGV Inc.; 603 Mission St.; Santa Cruz, CA 95060
Phone: (408) 427-4366 or (800) TGV-3440


> The metafile scheme is fine and clearly works well, but there is a MAJOR 
> problem if the little files start chewing up space and headers.  I feel very 
> strongly that if a package creates files (like .FDL... things) then it should 
> also clean them out if they become orphaned.  We use PacerShare as an 
> AppleShare file server on the VAX.  It uses these things too.  Recently we 
> ran out of file headers on the disk.  After writing a DCL script to clean out 
> the orphans we deleted about 5,500 orpahn files.  This is rediculous.

WOW!  Sounds like you hit a bug or a mis-design in the AppleShare file server.
I would wonder what the circumstances were that caused those AppleShare
files to stick around.  It turns out that the TGV NFS Client has the ability
to discover orphaned meta-files whenever the directory cache is refreshed.
We had many internal discussions on whether or not to "fix-up" the directory
by deleting those files and decided not to for the following reasons:
	1) Getting orphaned meta-files is apparently EXTREMELY rare.

	2) Silently going ahead and deleting a file someone might have
	   wanted was not a good thing.
If problems with orphaned meta-files show up then we would consider adding
something to the NFS client to scan for these files and, under close
supervision of a system manager, remove them.  Again -- this has not been
a problem and I don't expect it to become one.

(this, by the way, is quite a different argument than the argument we have
 been seeing from TWG that meta-files are BAD -- this is just that software
 that uses meta-files had better be VERY careful to keep proper track of them
 and get rid of ones that are of no more use).

> Please.  If you create files with your file NFS client s/w, please delete 
> them again if they are nolonger needed.  The files should be usable from both 
> the VAX and the UNIX sides and operations on either platform or another should 
> leave the file system lean, mean and efficient.  No excuses.
> 

Absolutely -- and this is the way that we do it.

David