Xref: utzoo comp.dcom.lans:6273 comp.sys.att:10643
Path: utzoo!attcan!uunet!cs.utexas.edu!sun-barr!olivea!orc!inews!iwarp.intel.com!gargoyle!chinet!les
From: les@chinet.chi.il.us (Leslie Mikesell)
Newsgroups: comp.dcom.lans,comp.sys.att
Subject: Re: Questions about StarGROUP Software
Message-ID: <1990Oct23.030821.9955@chinet.chi.il.us>
Date: 23 Oct 90 03:08:21 GMT
References: <1990Oct16.173745.18064@mccc.uucp> <15031@netcom.UUCP>
Organization: Chinet - Public Access UNIX
Lines: 16

I just ran into an interesting problem with the StarGroup DOS server
when used in combination with RFS.  It seems that a single server
process handles about 6 client PC's.  If one of those PC clients
has a file open on an RFS mounted filesystem, and the remote machine
goes down or otherwise forces an RFS disconnect, the StarGroup server
process dies, leaving 6 PC's in limbo.  If the process happens to be
the parent of the others, then the whole DOS server system goes out,
but new ones will start when the PC's reconnect.  If it isn't the
parent, only the PC's served by that process are affected, but when
they reconnect all the files they previously had open are locked.

Has anyone else seen this problem?  Is there a work-around?  Is it
fixed in 3.3?

Les Mikesell
  les@chinet.chi.il.us