HP3000-L Archives

February 1999, Week 4

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:
Jeff Woods <[log in to unmask]>
Reply To:
Jeff Woods <[log in to unmask]>
Date:
Fri, 26 Feb 1999 04:09:50 -0600
Content-Type:
text/plain
Parts/Attachments:
text/plain (39 lines)
At 2/25/99 11:34 PM -0700, Simonsen, Larry wrote:
>                                             why should I have to have a
>NT server and its ups so that I can cleanly shut my hp 3000 down?

You don't.  You can buy an HP UPS with software that can simply crash MPE (but
as I understand it not shutdown any applications, services, jobs, sessions or
anything else except halt the machine).  Should HP provide better handling than
that?  I think so and apparently some other folks do too.

But if you want to use a less expensive brand of UPS, who is responsible for
making it work?  If I buy an APC (or whatever non-HP brand) UPS like Ken
Sletten mentioned, should I expect HP to support it in MPE?  It would be nice,
but then I don't know if there are any standards for how such devices can be
connected to a host.  I suspect they all vary significantly.

If APC decides to support MPE with their software that is available for their
UPS, that would be great.  But I think the relative number of each host
platform APC is used with makes it highly doubtful that they will allocate the
resources to implement it.  Platforms with a relatively high visibility get the
finite resources of a major multi-platform vendor like APC.  That's why UNIX,
NT, Win9x and perhaps OS/2 and/or Linux get such support.  The number of folks
asking for MPE support likely just isn't sufficient to make it perceived as
significant to a company like APC in that marketplace.

On the other hand, if someone else implements it well and makes it freely
available to APC and/or their customers, perhaps they will provide a
distribution or even support for it.  Meanwhile, solutions which allow the
signal information that the power has failed to get from the UPS to the MPE
system have value whether they depend on other (admittedly less robust)
technology or not.

So the bottom line as I see it:  Do whatever is feasible to make it work as
robustly as possible given the tools and skills available.  For the moment on
this particular issue, I think that will require a significant degree of kludge
and custom implementation; but it still seems worthwhile to me.
--
Jeff Woods
[log in to unmask]  [PGP key available here via finger]

ATOM RSS1 RSS2