Frank, 

Get rid of that localhost. That sounds wrong.

Peter



On Tue, 23 Feb 1999, Frank Bennett wrote:

> Dear Peter,
> 
> On Mon, Feb 22, 1999 at 09:55:14AM -0500, Peter J. Braam wrote:
> 
> > It looks like you didn't make a root volume on the server.  Did you see
> > the Coda howto? 
> 
> I checked again, and unfortunately found that I had already carried
> through and made a root volume.  Just to be sure, I issued
> vice-killvolumes, then ran createvol_rep nagoya.law E0000100 /vicepa
> again.  I then shut down the server, which seemed to happen uneventfully,
> and then ran startserver again.
> 
> The server comes up fine, and reports in SrvLog that it has started. 
> SrvErr stayed clean. 
> 
> I have tried shutting down venus on the client machine, checking that
> /coda is unmounted, deleting the entire directory /usr/coda, and rerunning
> venus-setup 133.6.33.2 20000 several times, but always with the same
> result shown in the log files I sent to you yesterday.  Venus comes up
> without complaint, holds in memory for about 20-30 seconds or so (at a
> guess), then quietly goes away, leaving behind the log entries that I sent
> to you yesterday. 
> 
> There are three things I would like to check with you.  First, in
> /vice/hostname, I have just "localhost".  It seems to me that this should
> be okay, but I want to check to be sure.  Should this show the fully
> qualified domain name of the host? 
> 
> Second, I am trying to connect on this first session from a separate
> machine.  Could this be creating some sort of chicken and egg
> authorization problem on the server side?  I see that the first step after
> connecting is supposed to be to run clog, and I wonder whether the first
>  access has maybe got to be local, and that the first run of clog opens
> the server up to remote connections?  I would test this myself, but in the
> course of reallocating disk space on the server machine the compiler and
> all of the related utilities were lost, and it will be a few days before I
> have time to reinstall enough kit to bring up Venus on the server machine.
> 
> Third, when I sat at the console of the server and used Telnet to nudge
> the client machine to attempt a connection, I noticed that the server's
> disk with the log partition (on the server's main disk, but separate from
> the external disk carrying the Coda volumes) went into furious motion
> during the connection attempt, until the connection dropped.  I'm using
> the 130-meg data segment configuration.  Could the client or server simply
> be timing out?  Running codasrv --help, I see what look to be some
> timeout-related options.  Do you think I might get something back from
> tweaking one or more of these parameters? 
> 
> Looking forward to your next, I am
> -- 
> -x80
> Frank G Bennett, Jr         @@
> Faculty of Law, Nagoya Univ () email: bennett_at_nomolog.nagoya-u.ac.jp
> Tel: +81[(0)52]789-2239     () WWW:   http://rumple.soas.ac.uk/~bennett/
>