HP3000-L Archives

April 1999, Week 2

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 BARRETT <[log in to unmask]>
Reply To:
Steve BARRETT <[log in to unmask]>
Date:
Tue, 13 Apr 1999 13:32:00 +0000
Content-Type:
TEXT/PLAIN
Parts/Attachments:
TEXT/PLAIN (65 lines)
Tim Manns wrote:

I have been encountering problems very similar to ones listed in Tim
Herrin's posting.  However, mine have been mostly PCs running Reflection or
Minisoft, until this last weekend.  On the weekend, SOS/3000 (by Lund)
reported that the HP3000 had 90%+ idle time with about 20 users.  Even
though the 3000 was basically sitting idle, the response times got so bad
that a PC session would actually get disconnected because it timed out
waiting on the HP3000.  Terminals connected to the DTC72MX and DTC48 were
experiencing response times of 1 to 2 minutes.  All this with virtually no
system activity.  Seeing how we run SOS/3000 all the time, I had the ability
to review the log files generated by SOS from the weekend. The data showed
that we had high prompt response delays for all users.  However, the one
that was really strange was on the JSMAIN system process.  It had consistent
prompt responses of 30+ and even had a spike of 600+ in one 10 minute
interval!!

Any network/HP3000 gurus out here know what might be going on?  Anyone else
encountered any of these symptoms?  These problems all seem to have risen up
after we upgraded from MPE 5.0 to MPE 5.5 (powerpatch 6).  We also upgraded
our memory on the 3000 from 448MB to 764MB and installed new Fast Wide SCSI
disk drives. These upgrades were each done separately on different weekends
about 2 months ago.  We started noticing the performance problems about a
month to a month and a half ago and they have been getting progressively
worse.

Please feel free to email me if you have any ideas.

Thanks,
Tim Manns

--------------------------Message Separator-------------------

Tim,

We experienced a similar problem and determined (through HPRC)
that it was caused by "probe Proxy" being configured. Apparently,
if an appropriate proxy server isn't available on the network,
the HP3000 will beat itself to death trying to find one.

In NMMGR this is set via NS/NETEXPORT/GLOBAL.  Mine is set to 1,
0, 0.  The choices are:

    1   Network Directory
    2   Probe
    3   Probe Proxy
    0   Choice Omitted

Hope this helps.

Steve Barrett


    Steven P. Barrett
    Systems Analyst
    Fairfax County Public Library
    e-mail:  [log in to unmask]

    --- The opinions expressed here are mine alone. ---

    "Life is dear to every living thing, the worm that
    crawls upon the grownd will struggle for it."
    (from Solomon Northup's - Twelve Years a Slave, at -
    http://metalab.unc.edu/docsouth/northup/northup.html)

ATOM RSS1 RSS2