Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/5/84; site wucs.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxj!ihnp4!wucs!jdd
From: jdd@wucs.UUCP (John DeHart)
Newsgroups: net.eunice
Subject: EUNICE, TCP/IP installation problems
Message-ID: <447@wucs.UUCP>
Date: Mon, 29-Oct-84 12:23:19 EST
Article-I.D.: wucs.447
Posted: Mon Oct 29 12:23:19 1984
Date-Received: Tue, 30-Oct-84 01:45:04 EST
Distribution: net
Organization: Washington U. in St. Louis, CS Dept.
Lines: 53

We are having difficulties installing IP/TCP 1.28 on VMS V3.6 ( with
the 3.5 XGDRIVER) with Eunice 3.2. We have the shared Deuna version of 
IP/TCP and our DECNET works fine.
Everything works going out of VMS/EUNICE (ftp,rlogin,rsh, rcp,telnet,mail)
but rcp and rsh get some error messages when coming into VMS/EUNICE.
Mail has problems getting to VMS. Rlogin and ftp do not work at all. 
The error messages from rsh and rcp look like this

[EUNICE: AST ^Y setup failed, Status = %Xf4 - %SYSTEM-F-ILLIOFUNC illegal I/O]
[EUNICE: AST ^C setup failed, Status = %Xf4 - %SYSTEM-F-ILLIOFUNC illegal I/O]
[EUNICE: AST ^T setup failed, Status = %Xf4 - %SYSTEM-F-ILLIOFUNC illegal I/O]

These messages occur when rsh or rcp is used for the first couple of hours
after the network is restarted, after that the two commands work silently.
The commands (rcp, rsh) work even when these messages are received.

The last four lines of the mail messages look like this when trying to
mail to someone on the VMS machine.

% mail -v jdd@wucecv
... (normal messages)
451 (No such file or directory) Mail troubles (sndmsg balks), try later
>> QUIT
221 Bye received, Goodbye
jdd@wucecv ... Deferred: Bad file number
% 

rlogin and ftp on the other hand do not work at all. Ftp does the following

% ftp wucecv
connected to wucecv
%

Rlogin does similar things and sometimes gives messages like, "Broken pipe",
"rlogin: status 0".
I know that the daemons for rlogin and ftp get started, because they
show up on a SHOW/SYS (VMS) and on a ps -clax (EUNICE)

I have spent numerous hours on the phone with technical support from TWG
going over the installation procedure and everything looks fine.
They seem to have objections to the way we organized our VMS users.
We are supporting 1500-1600 students, so we set them up one level
deeper than normal (e.g DRB0:[USER.AAA9999]). We also use logical names
in defining our VMS users in AUTHORIZE (e.g. DEFINE U1 __DRB0:[USER.])
TWG has suggested that we use the physical name in AUTHORIZE and move
the users up a level, even though they admit it probably won't make
any difference.

Well there's our story. Anyone had similar problems? Any suggestions?
---
John DeHart                          ..!ihnp4!wucs!jdd
Box 1045 Washington Univ.          
St. Louis Mo. 63130 USA               (314) 889-6160