Minisoft odbc/BSQE table/hp1659
PAUL,GUY (HP-Boise,ex1)
You are correct that the HELLO.PE.VESOFT arms three handlers but each
handler for each logon takes an entry in the BSQE table. In addition to
VEsoft's 3 there are NSD's 6, plus CSM's 2, plus NBSPOOL's 3 for a total
of 14 armed handlers per session plus an additional number for jobs as well.
On HP1659 with 501 sessions and 126 jobs there are only 1450 entries left
in the table. On HP1660 with 765 sessions and 97 jobs there are 1300
entries left.
That leaves you about 100 logons left on HP1659 before filling up the
BSQE table.
Guy
-----Original Message-----
From: jsto...@mmm.com [mailto:jsto...@mmm.com]
Sent: Monday, February 18, 2002 18:53
To: guy_...@hp.com
Cc: US-IT-MP...@mmm.com; US-IT...@mmm.com; mechi...@mmm.com;
rrwa...@mmm.com; rtz...@mmm.com
Subject: Minisoft odbc/BSQE table/hp1659
Hello Guy -
when we consolidated Aberdeen we realized that their Minisoft ODBC job
requires Security 3000 to trap the HELLO command via the BACKG job.
Security 3000 uses AIF:PE's to accomplish this. I talked to Vesoft support
and they said that they arm 3 handlers, all at pin 1, and all bequeathed.
Given the issues with the BSQE table on hp1660, it would behoove us to keep
tabs on this table on hp1659. Can you take a couple of snapshots of the
BSQE table on hp1659 to see how many entries there are?
Also, a page from the Minisoft web site faq says that another option to
trapping the HELLO command via BACKG is to remove a segment from a Vesoft
SL that Security 3000 uses. That would let the Minisoft logons through
Security 3000 without problems. However, it also says Vesoft should be
contacted before doing this, because this may have an undesirable effect on
other functions of Security 3000. This would also mean that we would have
to remember to remove the segment every time we got a new SL with an
update.
The way we knew we filled the BSQE table up on hp1660 was that jobs started
aborted because of the run of jpenable.jpak.nsd was unsuccessful, and
JobPak picked up on the error message. We haven't seen that yet on hp1659
so I don't think we've hit the limit yet, but just in case, I've sent the
output of jpenable.jpak.nsd to $null in the epxeq.pub.nsd file, just like
we have it on hp1660.
Thanks Guy - Jeff
Jeff Stothart
Lawinger Consulting, Inc.
LCI Offices
106 Central Ave, Suite A
Osseo, MN. 55369-1243
Tel: (763) 425-9284
Office Fax: (763) 425-5483 Lotus Notes Fax: (651) 732-8212
jsto...@mmm.com
jeff.s...@lci-online.com
* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *