HP3000-L Archives

September 2001, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Ted Ashton <[log in to unmask]>
Reply To:
Ted Ashton <[log in to unmask]>
Date:
Mon, 3 Sep 2001 17:11:13 -0400
Content-Type:
text/plain
Parts/Attachments:
text/plain (137 lines)
Well, at long last, here's a wrap up of HPWorld from my perspective.  I was
fully intending to write a summary of Thursday that evening, but by the time
I got done redailing and reconnecting dozens of times for about 20 seconds
of connectivity each time, I was contemplating a move to Guatemala in hopes of
having a better a telephone system.

I started Thursday with Michael Hensley's "Test and Tune Your Disk Drives".
He's obviously done considerable testing and had some interesting conclusions.
It appears that simply adding more spindles is not necessarily going to help
(and in the case of serial reading can be a hinderance).  On the other hand,
adding controllers helped considerably as did getting faster drives.  Be sure,
however, to keep the newer drives cool--they tend to run very hot.  He advised
spending the extra money to get good enclosures.

The 9:00 session was Nora Denzel (Vice-President and General Manager, Network
Storage Solutions Organization) talking, of course, about storage.  She
reported that about 12 million TB of data had been created since the beginning
of time and that in the next 3 years, it is estimated that more data will be
created than has been created to this point.  She said that it used to be that
about 4% of the IT budget was spent on storage and that that is now running
anywhere from 7% to 20%.  In previous times, for each dollar spent on a server,
25 cents was spent on storage, 75 cents on the server.  In two years, she
expects that that will be exactly reversed--at the moment, about 50 cents is
spent each way.

The prediction is that within 3 years, a shop will have 10 times the data to
store that it has now and that the #1 problem in storage will be (and even now
to some extent is) staff shortage.

Storage has transitioned from "server-centric" -- directly connected storage
in a one-to-one relationship with the server to "storage-centric" -- storage
nodes on the network with many servers accessing the node.  Her prediction is
that the next move will be to "network-centric" storage -- many storage nodes
accessed by many different servers with a heterogeneous environment on both
sides.  In what she called "federated storage", the nodes would come from many
servers and would have to all be managed from a "single pane of glass".

She envisioned a new storage device introducing itself to the network,
registering its capabilities and thence being available for use.  HP hopes, of
course, to be a big part of the management of such networks.  OpenView, she
says, is currently being used to manage 70% of the world's networks, and more
non-HP than HP hardware is under its control.

The expectation is that the storage will become available to all operating
systems.  The file system will be a function of the storage, not of the server,
and any OS will be able to access the data at the record level.

Then she got off onto one of those wonderful, visionary ideas which make me a
bit nervous.  She presented the idea of "i-shadow".  When the time comes that
you personally own a TB of data, it's unlikely that you'll want to carry it
with you.  This mechanism would know where you are and make sure that your
data is always with you.  She pictured herself getting on the plane and
travelling to Chicago, walking onto stage and pointing here cell 'phone at
the screen and having the slides from her presentation just show up.  I'm not
sure I want the system constantly keeping track of where I am and gratuitiously
moving my data around for me.  Anywho.

Perhaps we will be able to carry it with us.  She said that atomic-resolution
storage gives promise of storage on the order of 500 TB/sq in.  Wow.  Also,
she put in a good mention of MPE during the talk in the context of IA-64 and
open systems.

Next on to the Strategic HP Management Roundtable.  Some good mention of MPE
there (with the as-expected leading of MPE in customer satisfaction).  Also the
only time during the conference that I heard the dreaded "three strategic
platforms" phrase.  HP, in my estimation, has improved considerably on that
front.

After a little time in the expo it was time for SIG IMAGE/SQL.  Ken has
reported well on that, so I'll let it be.  After that I went next door for
"ASPs As Invisible as Television" by Greg Stigers.  It was, unfortunately,
directly against "HP e3000 Server Directions" with Dave Snow, so the attendance
was, IIRC, 6.  I am unable at the moment to locate the notes from the
presentation, so I trust that Greg will correct me if I have misrememberd or
misunderstood.  IIRC, Greg dealt with the requirements necessary for making an
application into a application service--some of the technical needs but more
the non-technical--things like reasons for making the change and buzzword
compliance and suchlike.  It was a high-level overview sort of presentation.

Next to the e3000 Management Roundtable.  Considerably less tension in the
room this year :-).  The question about needing an "uncrippled" A-class box
received its answer from marketing (as expected).  They did considerable study
into the price-performance needs in the market and targeted the A-class and
N-class at specific spots.  They evidently met those spots fairly well, as
sales have been above expected.  The LDEV 1 limit issue is currently under
consideration at CSY.  I believe that the current exploration is partitioning
the drive, but I'd appreciate someone else setting me straight if need be.

The Fibre Channel answer is that it is work-in-progress, I believe.  They have
SCSI to FC--no converter required on the other end, and native FC is on its
way.  The answer to my question about Pascal is that the planned languages for
MPE on IA-64 are C, COBOL and Fortran.  As far as the "Applications on the HP
e3000" question, I don't remember it coming up so perhaps someone with a better
memory than mine could chime in.  The other question I remember was one from
SIGPrint about the Printronix printers.  CSY worked very hard to make sure that
folks knew that the LinJet printers were going to be sold by Printronix again,
but there were a few folks there who hadn't heard.  In case anyone amongst
this group hasn't heard, here's the letter:

  http://www.hp.com/products1/linejet/news_events/LINEJET001KU2.html

Thursday night was the HPWorld 2001 Party.  As there's been plenty said about
the volume, I'll say these things in favor: The pizza was good and they had an
artist who came in and did large paintings.  I was only there for the first
one--Groucho Marx, but I was impressed.

-- Friday --

I spent the first hour at "A World of Constraint" with Bruce Perens.  Wow.
He talked about the numerous constraints and invasions of privacy which we are
facing now or in the near future.  He was going to put the notes up on his
website (www.perens.com), but they aren't there yet.  In the end, he recommends
learning about the DMCA (the Digital Millenium Copyright Act) and writing (on
paper) to your representatives in Congress expressing your concerns about it.
There was much more to the presentation, but perhaps he'll get the notes up
on the site soon.

Next I picked up the second half of Jeff Vance's "MPE CI Programming".  As his
notes are available on Jazz, I'll move on to "Intermediate MPE/iX Performance".
Once again, I believe that the power point presentation was to be on the ITCC
website (www.craigs.com), but what's there at the moment is last year's
presentation (probably still good advice, in general :-).  After lunch I caught
some of the System Performance Panel and then started the journey home.  All
in all, a good conference.

Ted
--
Ted Ashton ([log in to unmask]), Info Sys, Southern Adventist University
          ==========================================================
Gott wurfelt nicht.
                                        -- Einstein, Albert (1879-1955)
          ==========================================================
         Deep thought to be found at http://www.southern.edu/~ashted

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

ATOM RSS1 RSS2