AAAREADME.1ST   -  Auto-trol Technology Corporation, DENVER, Colorado

This DECUS Spring 1988 Symposia TAPE submission contains the whole set
of management procedures used in managing an Ethernet VAX Cluster of
25 nodes in software development environment. 

The entire management philosophy may be briefly defined by:

 o CLUSTER must survive failure of several (BOOT) nodes
 o Node STARTUP must be the same regardless of BOOT node in use
 o STANDALONE Node boot must be similar (or same) as into CLUSTER 
 o Products should be separated form VMS, each in it's own directory
   tree, system startup only calls PRODUCT startup procedures

A document describing non-standard (sometimes unsupported) techniques
and CLUSTER management approaches used is enclosed in this top level 
directory.

Subdirectories contents:

 lavc$data:[LAVCCOM] - Procedures used in node startup/shutdown, user
		       login. Copy of systartup.com, syshutdown.com.
		       Cluster common files, such as SYSUAF.DAT have been
		       deleted for DECUS tape.

 lavc$data:[LAVCMGM] - Procedures used in cluster day-to-day management

 lavc$data:[CHICO]   - Database for the cluster "master" node CHICO

 lavc$data:[LISZT]   - Database for the workstation LISZT

 lavc$data:[GRP300]  - Databse for the user group 300


         	    D I S C L A I M E R

Author of this contribution does not claim any responsibility for
correctness and/or functionality of enclosed software, which is
provided primarily for reference purpose.  Procedures contained 
in this submissin have NOT been cleaned, they may contain significant 
ammount of installation specifc data.


			Martin Brunecky
			Advisory Engineer
			Auto-trol Technology Corporation
			12500 N. Washington St.
			DENVER, Colorado 80233
			(303) 252 - 2499