IUBio Biosequences .. Software .. Molbio soft .. Network News .. FTP

Move to Unix Platform

MICHAEL WEISE WEISE at bscr.uga.edu
Thu Apr 21 09:31:45 EST 1994


In <01HBEJ4NSDLK99EY1A at NIEHS.NIH.GOV> STAFFA at NIEHS.NIH.GOV writes:

> To my vast sorrow, there is a move from on high at this institution to
> put GCG on a Unix platform.  This machine is not yet bought. 
> The question  is:
> What platform, hardware and operating system,  would you recommend for
> GCG?

	Any specific recommendations would require more info re size of user
community, purchase budget, time frame for the migration, etc.  In our
particular situation, we've begun to look at a migration to Unix which we'd
consider doing 1-2 years from now.  Here are some of the things we've thought
about in relation to that kind of move.

	- We currently have DEC hardware configured as a Local Area VMS
	  cluster (LAVc).  Clustering under VMS is easy and solid.  It provides
	  good management tools and robust distributed batch job processing. 
	  It's therefore easy to have a system where most of the interactive
	  work is done on one 'large' platform (ours currently is a VAX
	  4000-100) and where batch processing is largely done on a set of
	  workstations (we have VAXstation 4000 90/60s) which are comparable in
	  power to the 'big' VAX.  This situation, in fact, is a major reason
	  for STAYING with VMS - at least for 'a while'.

	- All nodes in our LAVc are capable of running either VMS or OSF/1. 
	  Thus, it would be possible to switch to the OSF/1 flavor of Unix
	  without buying any new hardware.  (Presumably, there are licensing
	  issues, but those could hopefully be worked out throught DEC's
	  Educational Software Library program.) Since DEC has announced that it
	  will provide OSF/1 with cluster management and queueing functions, we
	  might conceivably move to a Unix OS without loosing any of the LAVc
	  niceties we've come to know and love.  (A DEC sales being has told us
	  that DEC is aiming to have OSF/1 with 'clustering' by the end of '95.)

	- New DEC hardware (those with the 64-bit 'Alpha' or AXP chip) can
	  pretty much 'plug-n-play' with an existing VAX-based LAVc.  This
	  provides a retirement path for 'old' VAXen and a path to take full
	  advantage of the new CPU architecture.  This seems really attractive
	  as OSF/1 AXP is already 64-bit in both OS _and_ file system (which
	  overcomes the current 2GB limit found in other flavors of Unix, quite
	  a benefit considering what sequence databases are doing).  Also, the
	  price/performace curves for the DEC AXPs suggest that they stack up
	  well compared to everything else currently on the market.
	  
	- Since GCG will soon have an Xwindows interface, and since VMS and
	  OSF/1 both provide a suite of DECwindows/Motif applications (editors,
	  a file/directory manger, mail...), providing users with the
	  capability to do X can ease an OS transition.  Ordinarily, users just
	  want to get some work done - they'd rather not invest time in
	  learning OS commands if they don't have to.  And, since point-n-click
	  VMS is likely to be much like point-n-click OSF/1, the training of
	  users with X might be minimal - even when the underlying OS is
	  changed.  (Of course, this tact has implications for networking and
	  usage of network bandwidth, but it's still good to 'plan big'.)

 
> I've noticed on INFO-GCG that people with Unix systems are always
> having some kind of trouble.  I've always thanked God for VMS and
> deleted these messages without paying them much mind. The questions: 
> What unique problems arise when running GCG on Unix?
> What sort of user-unfriendliness would I have to be warning my clients
> about when they are forced to make the transition? 

	The major problems will be moving to an OS which is case-sensitive and
which over-writes a file rather than creating another with a higher version
number.  Since users learn to live with PCs and Macs that don't do files with
version numbers, they should be able to get over loosing them when the move off
of VMS.  Case sensitivity is likely to be painful, especially around the time
of transition because of what filenames may look like when moved from a VMS to
a Unix system (unless the systems people are clever and do things to help).
However, since a number of VMS -> Unix transitions have already occurred,
there's presumably a body of knowledge in netland that should help ease things
for installations doing that in the future.

		MJW

   _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
 / Michael J. Weise, Ph.D.    \   Biological Sciences Computing Recource   \
(   weise at bscr.uga.edu         \   Dept.of Genetics  UGa, Athens  GA  30602 )
 \ _ _ _'Tis_only_me_speak'n._ _\_ _ _ _ _ _ _ (706) 542-1409_ _ _ _ _ _ _ /



More information about the Info-gcg mailing list

Send comments to us at biosci-help [At] net.bio.net