James,

  I'm not an Impromptu user so this may be way off the mark, but is there
a setting within the client (Impromptu) as to the amount of time it will wait
on the server to complete a transaction?  In many pieces of the client S/W
I've used there exists such an "escape" mechanism to keep the user from
doing a 'million row retrieval' :)
  If not, try turning some of the various trace options on to get a view of what
each piece thinks the other is doing during this time (like, is it returning the
tuples to the client, or is ODBCLink/SE waiting for Allbase to process the
query).  If it's returning the data to the client, maybe your running out of
memory and the message is bogus (that's a technical term for "in error" :)

BTW, the message you see is being generated by the ODBCLINK/SE
driver.  That is why it changed when you changed drivers.  This is also
what makes me suspect that Impromptu may be getting impatient.

Regards,
Michael L Gueterman
Easy Does It Technologies
email: [log in to unmask]
http://www.editcorp.com
voice: (888) 858-EDIT -or- (509) 943-5108
fax:   (509) 946-1170
--


----------
From:  James J. Kanihan[SMTP:[log in to unmask]]
Sent:  Monday, October 20, 1997 6:49 PM
To:  [log in to unmask]
Subject:  [HP3000-L] ODBCLink/SE Timeout Error

Hello MPE folks,

I've just discovered this amazingly informative newsgroup, so I hope
someone out there can help me.  So far, the HP Response Center sounds
like they're mostly guessing, and they haven't been able to help.  I'm
using Cognos Impromptu (version 4.01) to query an Allbase (version
A.G1.21) database, and I'm using the 32-bit ODBCLink/SE driver
(version E.55) under Windows NT Server 4.0, and MPE version 5.5.  If
the query tries to return a large (approx. 1 million) number of
records, it crashes with the error message "[ODBCLN32.DLL] Timeout on
data read" after about 15 minutes.  Cognos says that this is not their
message; it is being generated either by the database or ODBC driver.
<snip>