Known problems and workarounds

The "db2gcf -k" command fails on DB2 UDB Workgroup Server Edition

SQL1224 error from DRDA wrapper (AIX)

If a 32-bit DB2 Universal Database (UDB) server is run 7on an AIX system and an application running on the same system has more than 7one local database connection through the DRDA wrapper, then the application might get 7the following error:

SQL1822N  Unexpected error code "-1224" received from 
7data source "W3_SERVER2". 
7Associated text and tokens are 
7  func="DriverConnect" 
7  msg="SQL1224N  A database agent could not be started to 
7                 service a request, or was terminated as 
7                 a result of a database system shutdown 
7                 or a force command. "
7SQLSTATE=560BD

To avoid this error, put the following entry in the federated configuration 7file (instance_directory/cfg/db2dj.ini):

EXTSHM=ON

Alternately, you can catalog the local DB2 UDB database as being on a TCP/IP node. 7For example:

CATALOG TCPIP NODE my_node REMOTE my_host SERVER 123;
7CATALOG DB mydb AT NODE my_node;
7CREATE WRAPPER drda;
7CREATE SERVER my_server TYPE DB2/UDB VERSION 8 WRAPPER drda
7   AUTHORIZATION "my_id" PASSWORD "my_pw"
7   OPTIONS(ADD DBNAME 'MYDB');

Hotkeys not working in Microsoft Visual Studio .NET Framework 71.1

If your hotkeys are not working in Microsoft Visual Studio .NET Framework 71.1, you can download a hotfix from the Microsoft Web site. You can find the hotfix 7in the Microsoft Knowledge Base, article Q836745.

The simplified Chinese locale (AIX)

3AIX has changed the code set bound to the simplified Chinese locale Zh_CN 3on:

3
• AIX Version 5.1.0000.0011 or later
3
• AIX Version 5.1.0 with maintenance level 2 or later

The code set has been changed from GBK (code page 1386) to GB18030 (code 3page 5488 or 1392). Since DB2 Universal Database (UDB) for AIX supports the 3GBK code set natively and the GB18030 code set via Unicode, DB2 UDB will default 3the Zh_CN locale's code set to ISO 8859-1 (code page 819), and in some operations 3will also default the locale's territory to the United States (US).

To work around this limitation, you have two options:

• You can override the locale's code set from GB18030 to GBK and the territory from US to China (whose territory ID is CN and territory code is 86).
• You can use a different simplified Chinese locale.

If you choose to use the first option, issue the following commands:

db2set DB2CODEPAGE=1386
db2set DB2TERRITORY=86
db2 terminate
db2stop
db2start 

If you choose to use the second option, change your locale from Zh_CN to either ZH_CN or zh_CN. The ZH_CN locale's code set is Unicode (UTF-8), while the zh_CN locale's code set is eucCN (code page 1383).

5The simplified Chinese locale (Red Hat Linux)

5Red Hat Version 8 and later (including Red Hat Enterprise Linux [RHEL] versions 2.1 and 3) have changed the default code set for 5simplified Chinese from GBK (code page 1386) to GB18030 (code page 5488 or 51392).

3Since DB2 Universal Database (UDB) for Linux supports 3the GBK code set natively and the GB18030 code set via Unicode, DB2 UDB will default 3its code set to ISO 8859-1 (code page 819), and in some operations will also 3default its territory to the United States (US).

3To work around this limitation, you have two options:

3
• You can override the Red Hat default code set from GB18030 to GBK and 3the territory from US to China (whose territory ID is CN and territory code 3is 86).
3
• You can use a different simplified Chinese locale.

If you choose to use the first option, issue the following commands:

db2set DB2CODEPAGE=1386
3db2set DB2TERRITORY=86
3db2 terminate
3db2stop
3db2start 

If you choose to use the second option, issue any one of the 3following commands:

export LANG=zh_CN.gbk
3export LANG=zh_CN
3export LANG=zh_CN.utf8

where the code set associated with zh_CN is eucCN 3or code page 1383, and with zh_CN.utf8 is code page 1208.

Merant Driver Manager incompatibility (UNIX)

There are incompatibilities with Unicode support when the Merant Driver 5Manager accesses the DB2 ODBC driver on UNIX. These incompatibilities cause the 5Merant Driver Manager to use Unicode even if the application did not request 5Unicode use. This situation can lead to problems with components such as the 5Data Warehouse Center, Information Catalog Manager, and MQSI, which require 5the Merant Driver Manager to support non-IBM data sources. You can use an 5alternate DB2 ODBC driver library without Unicode support enabled until a permanent 5solution is available.

An alternative DB2 ODBC driver library without Unicode support enabled 5is included with DB2 Universal Database (UDB) Version 8.1 for AIX, 5HP-UX, and Solaris Operating Environment. To use this alternative library, 5you must create a copy of it, giving the copy the original DB2 ODBC driver 5library's name.

To switch to the non-Unicode ODBC library on AIX, HP-UX, or the Solaris Operating Environment, 5see the following instructions. Because this is a manual process, you must 5carry it out every time you update your product, including after the application 5of successive FixPaks or modification levels.

NFS APAR IY32512 - Unavailable threads (AIX)

AIX 5 NFS APAR IY32512 might cause the db2stop command to stop on systems with a large number of partitions.

On a server that is receiving large numbers of requests for blocking locks 2on files that are already locked, the lock daemon might become unresponsive. 2This situation occurs when all available locked threads are allocated to threads 2that are waiting for the locks to become available, so there is no thread 2available to pick up the work when the unlock request is made.

When this situation occurs the stopped nodes must be restarted. There 2is a DB2 Universal Database workaround for this situation by 2stopping the nodes one at a time using the NODENUM option of the db2stop command.

SQLFLAG (STD) precompiler option error

If the SQLFLAG(STD) precompile option is enabled, it will cause the following 4error: Abend C6 occurred while running Precompile program DSNHPC

Remove the SQLFLAG (STD) precompile option when using the Development Center 4to create SQL stored procedures to run on DB2 Universal Database for z/OS, Version 8.

DB2 Connect Custom Advisor

Despite being documented in the DB2 Connect User's 7Guide, the DB2 Connect Custom Advisor is no longer supported in Version 78.2.

Tools catalog database creation fails (HP-UX)

8

Problem

8

When installing DB2 Universal Database (UDB) Version 8.2 on 8HP-UX, the CREATE TOOLS CATALOG command fails with SQLCODE -22209 if it runs 8under a 32-bit instance and the DB2 Administration Server jdk_path configuration 8parameter points to the HP-UX SDK 1.4. The failure occurs because DB2 UDB Version 88.2 installs the HP-UX SDK 1.4 by default but the Version 8.2 32-bit instance 8requires the HP-UX SDK 1.3 to run the CREATE TOOLS CATALOG command successfully. 8

This failure can also occur when installing DB2 UDB Version 8.1 FixPak 7 if you manually 8updated the DB2 Administration Server jdk_path configuration parameter 8to point to the HP-UX SDK 1.4, or if you dropped and recreated the DB2 Administration 8Server (DAS). The failure occurs because, in either of these cases, the jdk_path configuration parameter changed to point to the HP-UX SDK 1.4.

8

A DB2 UDB Version 8.2 32-bit instance requires HP-UX SDK 1.3 to run successfully.

8

8

Workaround 1

8

Create the tools catalog under a 64-bit instance. 8

8

Workaround 2

8

Create the tools catalog under a 32-bit instance by performing the following 8steps: 8

8
1. Download the HP-UX SDK 1.3 from the HP-UX Web site: http://www.hp.com/products1/unix/java/
8
2. Install the HP-UX SDK 1.3.
8
3. Update the DB2 Administration Server jdk_path configuration parameter 8to point to the HP-UX SDK 1.3. For example: 8

   db2 update admin config using JDK_PATH /opt/java1.3

8
4. Restart the DB2 Administration Server. 8

   db2admin stop
   8db2admin start

8
5. Re-run the CREATE TOOLS CATALOG command under the 32-bit instance.

8

8

Displaying Indic characters in the DB2 GUI tools

If you have problems displaying Indic characters when using the DB2 GUI tools, you might not have the required fonts installed on your system.

DB2 Universal Database (UDB) has packaged the following IBM TrueType and OpenType proportional Indic language fonts for your use. You can find these fonts in the font directory on any of the following CDs:

7 7
• 7IBM Developer Kit, Java Technology Edition (64-bit) for AIX 75L7
7
• 7DB2 Embedded Application Server and applications 7(XML registry, Web Administration tools and Java distributed debugger) for your 7operating system7

These fonts are to be used only in conjunction with DB2 UDB. You cannot engage in the general or unrestricted sale or distribution of these fonts:

Detailed instructions on how to install the fonts and modify the font.properties file can be found in the Internationalization section of the IBM Development Kit for Java documentation.

In addition, the following Microsoft products come with Indic fonts that can be used with DB2 GUI tools:

• Microsoft Windows 2000 operating system
• Microsoft Windows XP operating system
• Microsoft Publisher
• Microsoft Office

GUI tools are not supported for zSeries servers (Linux)

With the exception of the DB2 Setup wizard, the GUI tools will not work on zSeries servers running the Linux operating system. This limitation includes any items normally launched from the Installation launchpad, such as the Quick Tour.

If you want to use the GUI tools with one of these systems, install the administrative tools on a client system with a different system configuration, and use this client to connect to your zSeries server.

Enclose DB2 Information Center search terms within quotation marks 7if they contain numbers

To get accurate search results in the DB2 Information Center you must enclose search 7terms that include numbers within quotation marks.

For example, if you search for the following term you will receive no results:

  1.4.1

However, if you enclose the term within quotation marks, you receive the 7appropriate results:

  "1.4.1"

A search for the following term will return extra topics:

  DB20000I

But a search on the following term works properly:

  "DB20000I"

Information Catalog Center log file not generated when importing tag 4language files

If an Information Catalog Center log file is not generated when you import 4tag language files to the Information Catalog Center, perform the following 4troubleshooting steps:

4

When running db2icmimport from a command line:

4

4

4
• If output files were not generated (.xml, .out, .err, .log), there is 4probably an error on the command line. Verify that the first five arguments, 4which are UserId, Password, Database, Catalog, and Tagfile, are correct. View 4the syntax by entering the db2icmimport command. 4If this does not solve the problem, modify db2icmimport to capture the output of the db2javit command 4by using the -g option to save the output to a file. For example: 4

  db2javit -j:com.ibm.db2.common.icm.tag.IcmImport -w: -i: 
  4   -o:"-Xmx128m -Xms32m" -g:"d:\temp\myimport.trc" ...

4
• If a log file was not generated, there is usually a parsing error. Look 4at the .xml file and the .out file. If you can, insert a ":COMMIT.CHKPID(DEBUG)" 4command at the beginning of the tag language file. This command generates 4debug report messages and checks the .xml and .out files for parsing errors.
4
• After parsing, errors should appear in the .log file. When the debug report 4is being generated, look at the .log file and the .out file for information.
4
• Always check the .err file to see if there is a runtime error.

4

4

4

When importing tag language files using the Information Catalog 4Center GUI:

4

4

4
• When you import tag language files using the GUI interface, no .out or 4.err files are generated.
4
• If a .log or .xml file is generated, try to debug by using those files.
4
• If a .log or .xml file is not generated or does not help, run the import 4process from a command line to get more information.

4

4

Binding Query Patroller packages

If the Query Patroller packages are not bound after applying a fixpak, 6a user without DBADM authority or proper Query Patroller privileges can encounter 6the following error when using the Query Patroller Center or Query Patroller 6command line:

SQL0001N - Binding or precompilation did not complete successfully.

If you are using the Query Patroller Center, the SQL0001N error is logged 6in qpdiag.log file. If you are using the Query Patroller command line, the 6SQL0001N is returned to the console

Auto-bind code exists to initiate automatic binding. However, the automatic 6binding fails when the connecting user does not have the necessary privileges 6to execute all statements in the Query Patroller packages. A symptom of this 6problem is missing folders in the Query Patroller Center.

To avoid this problem, the qpserver.lst packages should be bound manually 6by a user with DBADM authority or necessary privileges after applying a fixpak.

Unavailability of ports with Query Patroller (Windows)

Submitted queries in Query Patroller might receive SQL code -29007 when 9there are no more available ports on Windows XP or Windows 2003. 9The likelihood of this error increases with an increasing number of clients 9accessing Query Patroller.

   MaxUserPort=65534
9   TcpTimedWaitDelay=30 

Secure environments (Windows)

You might experience file permission problems if you are using DB2 Universal Database (UDB) on Windows and are not an administrator on the Windows system. 1If you receive an SQL1035N, SQL1652N, or SQL5005C error message, possible 1causes and workarounds are shown in the following information:

XML Extender sample programs renamed

Some XML Extender sample programs might have the same name as other installed 2programs. Accidentally invoking another program with the same name as an 2XML Extender sample program can damage your XML files. The following list 2shows the old XML Extender sample program names as well as new replacement 2program names that are less likely to cause conflicts. Make sure you use 2the new sample program names instead of the old ones to prevent damaging your 2XML files.

Using the new sample programs with the sample sqx files

The source code (.sqx files) for the executables listed previously are 2located in the samples\db2xml\c directory of your installation. 2The source files are still labeled with their old names. If you make changes 2to the source code, copy your newly compiled executables (with the old names) 2to the sqllib\bin directory.

On Windows platforms, you must make an additional copy, 2rename it with its new name above, and copy it to the bin directory. Both 2copies replace the existing files in the bin directory. For example, after 2compiling your new version of shred.exe, you need to make two copies and replace 2the files in the bin directory: one labeled shred.exe and the other renamed 2dxxshrd.exe.

On Linux and UNIX platforms, you need only to replace the file with the old 2name with your newly compiled version. If you create new executable files 2from these samples, you must copy the new files from the \SQLLIB\samples\db2xml\c\ directory into the \SQLLIB\bin\ directory, and then make 2an additional copy, renaming them according to the previous table.

Decomposing documents in XML Extender that contain non unique attribute 4and element names

You can now decompose documents that contain non-unique attributes or non-unique 4element names that map to different columns (of the same or different tables) 4without receiving the DXXQ045E error. The following is an example of an XML 4document with non-unique attributes and non-unique element names:

<Order ID="0001-6789">
4       <!-- Note: attribute name ID is non-unique -->
4       <Customer ID="1111">
4                    <Name>John Smith</Name>
4       </Customer>
4       <!-- Note: element name Name is non_unique -->
4       <Salesperson ID="1234"> 
4              <Name>Jane Doe</Name>
4       </Salesperson>
4       <OrderDetail>
4              <ItemNo>xxxx-xxxx</ItemNo>
4              <Quantity>2</Quantity>
4              <UnitPrice>12.50</UnitPrice>
4       </OrderDetail>
4       <OrderDetail>
4              <ItemNo>yyyy-yyyy</ItemNo>
4              <Quantity>4</Quantity>
4              <UnitPrice>24.99</UnitPrice>
4       </OrderDetail>       
4</Order>

The accompanying DAD, which maps the duplicate elements and 4attributes to different columns, looks like this:

<element_node name="Order">
4  <RDB_node>
4    <table name="order_tab" key="order_id"/>
4    <table name="detail_tab"/>
4    <condition>
4      order_tab.order_id=detail_tab.order_id
4    </condition>
4  </RDB_node>
4
4  <!--attribute ID duplicated below, but mapped to a different col-->
4  <attribute_node name="ID">
4    <RDB_node>
4      <table name="order_tab" />
4      <column name="order_id" type="char(9)"/>
4    </RDB_node>
4  </attribute_node>
4  
4  <element_node name="Customer">
4    <!--attribute ID duplicated above, but mapped to a different col-->
4    <attribute_node name="ID">
4      <RDB_node>
4        <table name="order_tab" />
4        <column name="cust_id" type="integer"/>
4      </RDB_node>
4    </attribute_node>
4
4    <!--element name duplicated below, but mapped to a different col-->
4    <element_node name="Name"> 
4      <text_node>
4        <RDB_node>
4          <table name="order_tab" />
4          <column name="cust_name" type="char(20)" />
4        </RDB_node>
4      </text_node>
4    </element_node>
4  </element_node>
4    
4  <element_node name="Salesperson">
4    <!--attribute ID duplicated above, but mapped to a different col-->
4    <attribute_node name="ID">
4      <RDB_node>
4        <table name="order_tab" />
4        <column name="salesp_id" type="integer"/>
4      </RDB_node>
4    </attribute_node>
4          
4    <!--element name duplicated above, but mapped to a different col--> 
4    <element_node name="Name"> 
4      <text_node>
4        <RDB_node>
4          <table name="order_tab" />
4          <column name="salesp_name" type="char(20)" />
4        </RDB_node>
4      </text_node>
4    </element_node>
4  </element_node>
4    
4  <element_node name="OrderDetail" multi_occurrence="YES">
4    <element_node name="ItemNo">
4      <text_node>
4        <RDB_node>
4          <table name="detail_tab" />
4          <column name="itemno" type="char(9)"/>
4        </RDB_node>
4      </text_node>
4    </element_node>
4    <element_node name="Quantity">
4      <text_node>
4        <RDB_node>
4          <table name="detail_tab" />
4          <column name="quantity" type="integer"/>
4        </RDB_node>
4      </text_node>
4    </element_node>
4    <element_node name="UnitPrice">
4      <text_node>
4        <RDB_node>detail_tab" />
4          <table name="detail_tab" />        
4          <column name="unit_price" type="decimal(7,2)"/>
4        </RDB_node>
4      </text_node>
4    </element_node>
4  </element_node>
4</element_node>

The contents of the tables would look like the following 4sample after the preceding document is decomposed:

ORDER _TAB:
4
4ORDER_ID       CUST_ID       CUST_NAME       SALESP_ID       SALESP_NAME
40001-6789      1111          John Smith      1234            Jane Doe
4
4DETAIL_TAB:
4
4ORDER_ID          ITEMNO         QUANTITY          UNIT_PRICE
40001-6789         xxxx-xxxx      2                 12.50
40001-6789         yyyy-yyyy      4                 24.99

Differences between SNA and TCP/IP when using DB2 Connect

When connecting to an OS/390 system using SNA, the host VTAM layer automatically 5 flows a commit when a new connection is made. The automatic commit allows 5the host side thread state to be inactive, and the thread immediately becomes 5 inactive.

However, when connecting to an OS/390 system using TCP/IP, there is no automatic 5 commit. The application itself must flow an explicit commit after the connection 5 to allow the thread to be inactive on the host. Without the explicit commit, 5the thread is subject to an idle thread time out.

The suggested workaround is to rewrite the application so that it will 5perform an explicit commit if the connection goes idle after the connection.