IBM Support

Collecting data for InfoSphere Federation Server Crash Issues

Question & Answer


Question

What information should be collected when I experience a crash problem with my software? Collecting data for crash issues. Collecting this information before calling IBM support will help you understand the problem and save time analyzing the data.

Answer

Provide the following output for looping or hanging problems
1. Exact name of product installed by issue this command and operating system command prompt:
db2licm -l

2. Current version of DB2 engine level. Enter the following from an operating system command prompt:
   db2level

3. The operating system type, and release, where Federation Server is installed.

    On UNIX/Linux, enter the following from an operating system command prompt:
       uname -a
      For AIX, also run below command:
      oslevel -s

      For Linux, provide the vendor/flavor/release too
      - On SuSE Linux, enter:
          cat /etc/SuSE-release
      - On RedHat Linux, enter:  
          cat /etc/redhat-release

    On Windows, right click on "My Computer", then select the Properties option.

4. What's the topology of your Federation Solution? Is Federation connecting to more than one data sources ? What are they? Provide data source client and server exact release and version.
Examples commands:
    DB2 for LUW
    Enter the following from an operating system command prompt:
    db2level

    Informix
    Enter following from an operating system command prompt:
    cd $INFORMIXDIR/bin
    ./ifx_getversion clientsdk
    ./ifx_getversion libasf.a
    ./ifx_getversion libasf.so

    Oracle
    Enter the following from an operating system command prompt:
    cd $ORACLE_HOME/bin
    ./sqlplus

    Microsoft SQL Server
    Issue the following SQL statement (from the Microsoft Query Analyzer):
    select @@version

    IBM Branded DataDirect ODBC driver
    Go to directory <IBM_branded_odbc_install_dir>/bin and execute below command:
    ddtestlib <libname>

    For example:
    i) Get ODBC driver manager version:
    ddtestlib <IBM_branded_odbc_install_dir>/lib/odbc.so

    ii) Get Microsoft SQL Server ODBC driver version:
    ddtestlib <IBM_branded_odbc_install_dir>/lib/VMmsss23.so


    Sybase
    sybase client version:
    Go to directory $SYBASE/$SYBASE_OCS/bin and execute below command:
    isql -v

    Sybase server version:
    Issue the following SQL statement after connecting database using isql command:
    select @@version
    go

    Teradata
    Start DWB
    issue get version command

5. Have there been any recent changes to your environment (hardware, operating system, software version)?

6. Has this problem occurred before? If yes, do you have a previous PMR number?

7. How often does this problem occur?

8. Can the problem be reproduced? If so, provide db2 trace and exact details of the steps performed when problem encountered.
      # turn on tracing and trace all components
    • db2trc on -f trace.trc

    <Run command that recreate the error>
      NOTE: DB2 will write information into the trace file right before the crash so you can format it after bringing DB2 instance back again.

      # turn off tracing
    • db2trc off
      # format the trace flow diagram into file trace.flw
    • db2trc flw trace.trc trace.flw
      # format the trace data into file trace.fmt
    • db2trc fmt trace.trc trace.fmt
      # format the trace data into file trace.fmc if DRDA wrapper
    • db2trc fmt -c trace.trc trace.fmc

    ii) DDLs for the remote tables on datasource

    iii) Federation objects setup:
    db2look -d <dbname> -e -o db2look.out

    iiii) Query or any operations that triggered the trap
If not, describe the details of your application's implementation, such as any simultaneous DDL,insert,update,delete or query on nicknames.

9. Double check log files generated when crash happened
    i) For DB2 engine(db2sysc) crash, there should be FODC_Trap directory generated like this:
    <instance path>/sqllib/db2dump/FODC_Trap_<TIMESTAMP>

    Double check your db2diag.log, it should includes all the messages generated from crash time back to the time DB2 ran well.

    ii) For db2fmp crash, there should be trap files for db2fmp process:
    <instance path>/sqllib/db2dump/<db2fmp process ID>.<db2fmp thread ID>.000.app_stack.txt

    iii) On Windows, use the following command to format above trap files before sending to IBM Support. For example:
    db2xprt 3012.5308.trap.bin 3012.5308.trap.bin.txt
10. Collect db2support
    db2support <output_path> -d <db name> -cl 0


11. Is this a production, development, or test environment?

12. What is the business impact of this problem?
Examples:
    a. Cannot roll out a new application
    b. Cannot monitor in-store security systems
    c. Losing money due to applications not talking to each other

13. Zip, or compress, all files and rename with PMR number as prefix before submitting ; example 12345.678.999.files.zip.


Submitting information to IBM Support

Once you have collected your information, you can begin Problem Determination through the product Support web page, or simply submit the diagnostic information to IBM support. Use the URL below for submitting information to IBM Support.
http://www.ecurep.ibm.com/app/upload

For a listing of all technotes, downloads, and educational materials specific to the InfoSphere Federation Server component, search the product support site at the following link : http://www-01.ibm.com/software/data/infosphere/support/federation-server/

[{"Product":{"code":"SS2K5T","label":"InfoSphere Federation Server"},"Business Unit":{"code":"BU001","label":"Analytics Private Cloud"},"Component":"General Issues","Platform":[{"code":"PF002","label":"AIX"},{"code":"PF010","label":"HP-UX"},{"code":"PF016","label":"Linux"},{"code":"PF027","label":"Solaris"},{"code":"PF033","label":"Windows"}],"Version":"9.7;10.1;10.5","Edition":"All Editions"}]

Document Information

Modified date:
16 June 2018

UID

swg21442088