HP3000-L Archives

November 1995, 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:
Steve Cole <[log in to unmask]>
Reply To:
Date:
Sun, 5 Nov 1995 00:22:31 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (44 lines)
In a message dated 95-11-04 22:54:11 EST, [log in to unmask] (Richard
Gambrell) writes:
 
>Steve, what (these days) *is* a "large" HP3000 system? I used to think that
>anything larger that a 949 was a "large" system (more than 100 users), then
>it seems like it was a system with a "6" or larger as the middle digit.
>Today, a case could be made that the only large systems are multi-cpu
systems
>or those with two cabinets (and I'm not talking about DTC cabinets).  What
>did you mean my "large" and is a 967 "large"?
>
>In fact, why would you not recommend it for smaller systems? I recall a
>program that used to hog the cpu whenever it would run (interactive queue)
>and
>I wished for an easy way to limit this effect without rearranging all of the
>queues and other processes - wouldn't the workload manager assist with a
>problem like that even on a small system?
>
>
 
With the rapid change in technology with HP today there is a large variety
of HP3000s out today.  The variety includes 950s, 955s, 925s, 922s and the
upgrades to these systems including the 955s through 958s.  What I
consider to be midrange of the older CIO technology is the 960 and 958
systems.  When looking at the newer NIO technology then 957 would
be midrange.
 
In regards to using the Workload Manager on smaller systems I agree that
it would work.  The problem with the Workload Manager on smaller systems
is that the interrupt scheme is not as grandular as it was on previous
releases.  This means that the slower the system the less control you
have.  I tested the Workload Manager on systems from the 917 to the 992.
The results of the testing was that the actual results received on a 917
was not as accurate as on a 960.  As the system got larger or if
multi-processors were involved the Workload Manager worked more
accurately.
 
I feel the Workload Manager is a valuable tool and I recommend it
when additional control is required.  However, the results you actually
receive may be related to the actual performance of the CPU.
 
Steve Cole
Outer Banks Solutions

ATOM RSS1 RSS2