| Author |
Message |
Martin Cronje
Guest
|
 Posted:
Fri Nov 26, 2004 9:15 pm Post subject:
Host Integration Server 2004: MSDB2 .Net Provider load probl |
|
|
I am building an application which has to execute SQL INSERT statement
on a OS/390 DB2 Database, using DTC as I need to run a MS SQL Server
update in the transaction as well.
We are running into an error with the updates to DB2 using the MSDB2
..Net Class library, where on both the Mainframe and Windows
environment the other party seems to close the network connection.
This error happens intermittently and seems to be connected to the
size of the insert statements to the DB2 database:
· 15 Fields in insert statement with approximately 200-300 inserts
· 60 Fields in insert statement with approximately 50-100 inserts
The transactions are batched into groups of 20 (2 Second expiry on
batch population) and each batch runs in its own thread.
The moment I increase the load significantly the error causes the
mainframe's listener (DDF) to ABEND. The listener has to be manually
restarted when this happens.
Something else I picked up was that the connection stays active for a
period of 30 seconds on the mainframe after the transaction was
committed (This may be part of the problem?)
The error being logged to the event log states:
The network connection was terminated because the host failed to send
any data. SQLSTATE: 08S01, SQLCODE: -605
On the mainframe an error is logged with the following information:
14.01.30 STC06377 DSNL032I -XXXX DSNLIRTR DRDA EXCEPTION CONDITION
IN
REQUEST FROM REQUESTOR LOCATION=10.3.8.26 FOR
THREAD WITH
LUWID=GA03081A.G8D2.BBF8D257BB02
REASON=00D3101B
ERROR ID=DSNLIRTR0004
IFCID=0192
SEE TRACE RECORD WITH IFCID SEQUENCE
NUMBER=00000000
14.01.38 STC06377 DSNL033I -XXXX DSNLILNR
DDF TERMINATION SCHEDULED DUE TO DETECTION OF
ABEND 0C4
REASON 00000000
UDL file:
[oledb]
; Everything after this line is an OLE DB initstring
Provider=DB2OLEDB;Password=password;Persist Security Info=True;User
ID=devUser;Initial Catalog=XXXX;APPC Security Type=PROGRAM;Network
Transport Library=TCPIP;Host CCSID=37;PC Code Page=1252;Network
Address=MVSHost1;Network Port=446;Package Collection=DB1G;Default
Schema=DB1G;DBMS Platform=DB2/MVS;Process Binary as
Character=False;Connection Pooling=True;Units of Work=DUW
Does anyone have an idea why this may happen? Any help will be greatly
appreciated!
Regards,
Martin |
|
| Back to top |
|
 |
Jeremy Remlinger [MSFT]
Guest
|
| Posted:
Thu Dec 02, 2004 2:54 am Post subject:
RE: Host Integration Server 2004: MSDB2 .Net Provider load p |
|
|
Martin,
Given your description, you may want to focus on the host side here. Since
the host application is giving an ABEND, we need to know why. This may
require some debugging on the host side. I would discuss this furhter with
your host administrators.
If you are connecting via TCP/IP, you may want to take a network sniff and
see what is happening prior to this ABEND. Look for reset TCP sessions,
etc.
#############################
#Jeremy "Rem" Remlinger, MSCE
#SNA/HIS Engineer
#Microsoft
#
#This posting is provided
#'AS IS' with no warranties,
#and confers no rights.
#© 2004 Microsoft Corporation
#All rights reserved.
#############################
--------------------
From: martincronje@hotmail.com (Martin Cronje)
Newsgroups: microsoft.public.hiserver.general
Subject: Host Integration Server 2004: MSDB2 .Net Provider load problem.
Date: 26 Nov 2004 07:15:00 -0800
Organization: http://groups.google.com
Lines: 67
Message-ID: <e483705b.0411260715.6e654295@posting.google.com>
NNTP-Posting-Host: 196.41.200.170
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Trace: posting.google.com 1101482101 29789 127.0.0.1 (26 Nov 2004
15:15:01 GMT)
X-Complaints-To: groups-abuse@google.com
NNTP-Posting-Date: Fri, 26 Nov 2004 15:15:01 +0000 (UTC)
Path:
cpmsftngxa10.phx.gbl!TK2MSFTFEED02.phx.gbl!TK2MSFTNGP08.phx.gbl!news-out.cwi
x.com!newsfeed.cwix.com!news.maxwell.syr.edu!postnews.google.com!not-for-mai
l
Xref: cpmsftngxa10.phx.gbl microsoft.public.hiserver.general:4464
X-Tomcat-NG: microsoft.public.hiserver.general
I am building an application which has to execute SQL INSERT statement
on a OS/390 DB2 Database, using DTC as I need to run a MS SQL Server
update in the transaction as well.
We are running into an error with the updates to DB2 using the MSDB2
.Net Class library, where on both the Mainframe and Windows
environment the other party seems to close the network connection.
This error happens intermittently and seems to be connected to the
size of the insert statements to the DB2 database:
· 15 Fields in insert statement with approximately 200-300 inserts
· 60 Fields in insert statement with approximately 50-100 inserts
The transactions are batched into groups of 20 (2 Second expiry on
batch population) and each batch runs in its own thread.
The moment I increase the load significantly the error causes the
mainframe's listener (DDF) to ABEND. The listener has to be manually
restarted when this happens.
Something else I picked up was that the connection stays active for a
period of 30 seconds on the mainframe after the transaction was
committed (This may be part of the problem?)
The error being logged to the event log states:
The network connection was terminated because the host failed to send
any data. SQLSTATE: 08S01, SQLCODE: -605
On the mainframe an error is logged with the following information:
14.01.30 STC06377 DSNL032I -XXXX DSNLIRTR DRDA EXCEPTION CONDITION
IN
REQUEST FROM REQUESTOR LOCATION=10.3.8.26 FOR
THREAD WITH
LUWID=GA03081A.G8D2.BBF8D257BB02
REASON=00D3101B
ERROR ID=DSNLIRTR0004
IFCID=0192
SEE TRACE RECORD WITH IFCID SEQUENCE
NUMBER=00000000
14.01.38 STC06377 DSNL033I -XXXX DSNLILNR
DDF TERMINATION SCHEDULED DUE TO DETECTION OF
ABEND 0C4
REASON 00000000
UDL file:
[oledb]
; Everything after this line is an OLE DB initstring
Provider=DB2OLEDB;Password=password;Persist Security Info=True;User
ID=devUser;Initial Catalog=XXXX;APPC Security Type=PROGRAM;Network
Transport Library=TCPIP;Host CCSID=37;PC Code Page=1252;Network
Address=MVSHost1;Network Port=446;Package Collection=DB1G;Default
Schema=DB1G;DBMS Platform=DB2/MVS;Process Binary as
Character=False;Connection Pooling=True;Units of Work=DUW
Does anyone have an idea why this may happen? Any help will be greatly
appreciated!
Regards,
Martin |
|
| Back to top |
|
|
Charles Ezzell (MSFT)
Guest
|
| Posted:
Thu Dec 02, 2004 7:06 pm Post subject:
Re: Host Integration Server 2004: MSDB2 .Net Provider load p |
|
|
The SQL Code is "The network connection was terminated because the host
failed to send any data. "
Internally, we have only seen this error one time, and that was with
parameterized queries with FLOAT failing on AIX. The test we were using was
creating a binding with "DBTYPE_NUMERIC" and a precision of 15, but the
float values have more than 15 digits of precision so the test was failing.
Changing the binding to DBTYPE_R8 solved the problem.
As a side note: We have a 128 command limitation. In trying to set this to a
larger value, we noticed performance degradation on some host systems, so
the decision was made to impose this limit.
Charles
"Martin Cronje" <martincronje@hotmail.com> wrote in message
news:e483705b.0411260715.6e654295@posting.google.com...
| Quote: | I am building an application which has to execute SQL INSERT statement
on a OS/390 DB2 Database, using DTC as I need to run a MS SQL Server
update in the transaction as well.
We are running into an error with the updates to DB2 using the MSDB2
.Net Class library, where on both the Mainframe and Windows
environment the other party seems to close the network connection.
This error happens intermittently and seems to be connected to the
size of the insert statements to the DB2 database:
· 15 Fields in insert statement with approximately 200-300 inserts
· 60 Fields in insert statement with approximately 50-100 inserts
The transactions are batched into groups of 20 (2 Second expiry on
batch population) and each batch runs in its own thread.
The moment I increase the load significantly the error causes the
mainframe's listener (DDF) to ABEND. The listener has to be manually
restarted when this happens.
Something else I picked up was that the connection stays active for a
period of 30 seconds on the mainframe after the transaction was
committed (This may be part of the problem?)
The error being logged to the event log states:
The network connection was terminated because the host failed to send
any data. SQLSTATE: 08S01, SQLCODE: -605
On the mainframe an error is logged with the following information:
14.01.30 STC06377 DSNL032I -XXXX DSNLIRTR DRDA EXCEPTION CONDITION
IN
REQUEST FROM REQUESTOR LOCATION=10.3.8.26 FOR
THREAD WITH
LUWID=GA03081A.G8D2.BBF8D257BB02
REASON=00D3101B
ERROR ID=DSNLIRTR0004
IFCID=0192
SEE TRACE RECORD WITH IFCID SEQUENCE
NUMBER=00000000
14.01.38 STC06377 DSNL033I -XXXX DSNLILNR
DDF TERMINATION SCHEDULED DUE TO DETECTION OF
ABEND 0C4
REASON 00000000
UDL file:
[oledb]
; Everything after this line is an OLE DB initstring
Provider=DB2OLEDB;Password=password;Persist Security Info=True;User
ID=devUser;Initial Catalog=XXXX;APPC Security Type=PROGRAM;Network
Transport Library=TCPIP;Host CCSID=37;PC Code Page=1252;Network
Address=MVSHost1;Network Port=446;Package Collection=DB1G;Default
Schema=DB1G;DBMS Platform=DB2/MVS;Process Binary as
Character=False;Connection Pooling=True;Units of Work=DUW
Does anyone have an idea why this may happen? Any help will be greatly
appreciated!
Regards,
Martin |
|
|
| Back to top |
|
|
|
|
|
|
FAQ
Memberlist
Register
Profile
Log in to check your private messages
Log in