Collecting data for instance management problems

If you are experiencing problems while performing instance management and you cannot determine the cause of the problem, collect diagnostic data that either you or IBM Software Support can use to diagnose and resolve the problem.

These steps are only for situations where you can re-create the problem and you are using Db2® on Linux® or UNIX.

To collect diagnostic data for instance management problems:

  1. Repeat the failing command with tracing or debug mode enabled. Example commands:
    db2setup -t trace.out
db2icrt -d INSTNAME
db2idrop INSTNAME -d
db2iupgrade -d INSTNAME
db2iupdt -d INSTNAME
  2. Locate the diagnostic files. More than one file might be present, so compare the timestamps to ensure that you are obtaining all of the appropriate files.

    The output will be found in the /tmp directory by default.

    Example file names are: db2icrt.log.PID, db2idrop.log.PID, db2iupgrade.log.PID, and db2iupdt.log.PID, where PID is the process ID.

  3. Provide the diagnostic file(s) to IBM Software Support.

If the problem is that the db2start or START DATABASE MANAGER command is failing, look for a file named db2start.timestamp.log in the insthome/sqllib/log directory, where insthome is the home directory for the instance owner. Likewise if the problem is that the db2stop or STOP DATABASE MANAGER command is failing, look for a file named db2stop.timestamp.log. These files will only be found if the database manager did not respond to the command within the amount of time specified in the start_stop_time database manager configuration parameter.