Reviewing the Siebel Upgrade Log Files

This chapter describes the upgrade log files that the Siebel Upgrade Wizard produces on the midtier during the upgrade file generation process. It also describes how to check the status of staging and target upgrade jobs on the z/OS host using the z/OS Siebel job logs. This chapter includes the following topics:

About the Siebel Upgrade Log Files

The Upgrade Wizard writes logs that provide detailed information on the upgrade processes and they also list all errors. The Upgrade Wizard writes the logs for a process to the following directory by default:

Windows: SIEBEL_ROOT\LOG\process

UNIX: $SIEBEL_ROOT/log/process

where process is the name of the upgrade process you have run, for example, upgrep_dev_782 or prepare_for_production_upgrade.

The process directory contains the following subdirectories:

Output. Directory containing the Upgrade Wizard log files
State. Directory containing the state.log file

The output and state directories are automatically archived on subsequent runs of a process that completes successfully. (The names of subsequent log directories are appended with _1, _2, and so on.) To preserve disk space, periodically delete or save log directories to another location.

Note: You can select a different log directory from the Log Output Directory screen on the Database Configuration Wizard.

About the State Log File

Each upgrade process consists of a series of steps, each of which must complete successfully. If the Upgrade Wizard cannot complete a step, it marks the step as incomplete in the state.log file and exits. The state.log file is located in SIEBEL_ROOT\LOG\process\state (Windows) or SIEBEL_ROOT/LOG/process/state (UNIX).

You must correct the error and then run the Upgrade Wizard again. When you rerun the Upgrade Wizard, it refers to the state log and resumes at the incomplete step that contained the error.

About Process Log Files

You can identify errors you encounter during an upgrade by reviewing the log file named UpgWiz.log (Windows) or srvrupgwiz1.log (UNIX) in the SIEBEL_ROOT\LOG\process\output directory (Windows) or the SIEBEL_ROOT/LOG/process/output directory (UNIX).

The name of the log file increments for subsequent log files that are created if the Siebel Upgrade Wizard encounters a problem and you run the Siebel Upgrade Wizard again.

Review the end of the log file for details about the latest failure. If the step that failed was not a native SQL step (which would be listed in the log file), then it occurred as part of an external utility. You can review the relevant log file, which is identified by the/L parameter.

How to Determine if the Upgrade Process Completed Successfully

If the status of all the steps in the state.log is Complete, the upgrade process completed successfully.

If the status of any step is Incomplete, the upgrade process did not complete successfully. You must identify the error and correct it before resuming the upgrade.

Note: In some cases, the Upgrade Wizard can complete a step even though the step contains unacceptable errors. You must verify that all steps do not contain unacceptable errors, even those with a status of Complete.

Use the following process to identify errors:

Resolve errors for steps identified with a Status of Incomplete in the state.log file.
Review all the steps with a status of Complete in the state.log file. If any contain unacceptable errors, resolve these errors. See Reviewing Siebel Upgrade Log Files for Errors for information on identifying unacceptable upgrade errors.
Restart the Upgrade Wizard, or, if necessary, restore the database and rerun the upgrade process.

If you have any questions regarding how to resolve errors, create a service request (SR) on My Oracle Support. You can log service requests by accessing My Oracle Support (Service Request tab), or by using your existing phone support numbers to contact Oracle Global Customer Support.

Log Files That Can Be Ignored

If the upgrade completed successfully, there are several log files that you can safely ignore:

Windows: sw_cfg_xxx.log and siebel.log
UNIX: srvrupgwiz_*.log and siebel_*.log. For example, srvrupgwiz_001.log, and srvrupqwiz1_02.log
Any other log file that existed before the start of the upgrade

Caution: UNIX Only : The log file srvrupgwiz_001.log is a different file than srvrupgwiz1.log. Do not ignore log files named srvrupgwiz1.log, srvrupgwiz1_01.log and so on.

Reviewing Siebel Upgrade Log Files for Errors

Upgrades: All upgrades.

Environments: All environments.

This topic is part of an upgrade process. See How to Perform a Siebel Database Upgrade.

Review the logs created when you run the Siebel Upgrade Wizard to verify that the upgrade process completed correctly and to identify errors that must be resolved. The log files might include errors that are expected and benign. You must compare any error messages found in the log files to a list of acceptable error messages, and correct any unacceptable errors.

Complete the following procedure to manually review log files for unacceptable errors.

To manually review the log files for unacceptable errors

Review the state.log file to see at what step the upgrade failed. This step can be traced back to the driver file. The state.log file is located in the following directory:
Windows: SIEBEL_ROOT\LOG\process\state
UNIX: $SIEBEL_ROOT/LOG/process/state
Print the errors file. The errors file lists the benign and expected errors you might find in the log files; you can ignore these errors. The errors file is located in the installation subdirectory:
Windows: DBSRVR_ROOT\DB2390\errors.rtf or errors.htm
UNIX: DBSRVR_ROOT/DB2390/errors.txt
Sort the log files in the following directory by date.
Windows: SIEBEL_ROOT\LOG\process\output
UNIX: $SIEBEL_ROOT/LOG/process/output
Open each log file, starting with the earliest, and search for errors. Starting with the earliest log file can shorten your research time.
Log files are identified by the .log extension. Errors are either tagged with the word error or enclosed in square brackets [...].
For each error found, compare the error description against the list of acceptable errors documented in the errors file.
The log files generated by the Siebel Upgrade Wizard (for example srvrupgwiz1.log) appear in the errors file as upgwiz1.log, upgwiz2.log, incrementing for additional log files. Identify errors as follows:
- If you find the error in the errors file, it is acceptable and no action is required. Continue to review the errors found in the log file.
- If an error appears multiple times in a log file, but only one occurrence of that error appears in the errors file, all errors of that type are acceptable and no action is required. Continue to review the errors found in the log file.
- If a log file is not listed in the errors file, there are no acceptable error messages for that log file. You must correct the condition that caused the error before you rerun the Siebel Upgrade Wizard.
- If you find an error that is not listed in the errors file, it is unacceptable. You must correct the condition that caused the error before you rerun the Siebel Upgrade Wizard.
  To help resolve unacceptable errors, view 477324.1 (Article ID) on My Oracle Support. This document was formerly published as Siebel Troubleshooting Steps 21. Common errors and how to resolve them are listed in this document. If the error is not listed or you cannot resolve it, create a service request (SR) on My Oracle Support. Do not proceed with the upgrade.
Repeat step 5 for each log file.

Although unacceptable errors are rarely encountered, this review is critical. Certain errors, such as a failure to create indexes, can result in performance problems or anomalous behavior in Siebel Business Applications.

Manually Archiving Upgrade Log Files

Upgrades: All upgrades.

Environments: All environments.

This topic is part of an upgrade process. See How to Perform a Siebel Database Upgrade.

After a successful installation and upgrade, you must manually save and archive the log files located in the SIEBEL_ROOT/LOG/process (Windows) directory.

By default, only nine (9) upgrade log files are retained for subsequent retries of the Siebel Upgrade Wizard. After nine log files have been created, when the Siebel Upgrade Wizard is rerun, it overwrites log files beginning with the earliest one created and recycles the rest as necessary. (This does not apply to the state.log file.)

The number of log files retained can be increased by resetting the siebel_log_archive environment variable, for example, set the variable to 20 to retain twenty (20) log files.

Viewing the Siebel Job Log Status

Upgrades: All upgrades.

Environments: All environments.

This topic is part of an upgrade process. See How to Perform a Siebel Database Upgrade.

Whether you are using Siebel-scheduling or vendor-scheduling to run your upgrade jobs, you can query the Siebel job log for the staging and target upgrade processes by completing the following procedure.

Note: To view job status, you must have installed DSN REXX.

To view the Siebel job log status

If you are not on the Siebel In-Place Upgrade Main Menu, enter the following command:
```
EXEC ’DSNHLQ.SIEBEL.EXEC’
```
Select one of the following options, and press Enter:
- Option 5: Staging Database Joblog (SBLLOG S).
  The Staging Joblog Query panel displays.
- Option 6: Target Database Joblog (SBLLOG T).
  The Target Joblog Query panel displays.
The panel ID of both the staging and target joblog panels is SBLLOGP.
A list of successful and failed jobs appears.
Next to the label, List By Job Status, enter 1 to list failed jobs and enter 2 to list jobs that have not yet been run.
The list displays 250 lines only. The Unload, Load and Index Rebuild jobs have more than 250 jobs, so you must query using another option or by specific or partial job name.
Press PF3 when you are finished viewing the log.
The Siebel In-Place Upgrade Main Menu for your upgrade path appears.

Running SQL in Siebel Logs

By using SPUFI or the command line, you can construct SQL queries to run against the staging or the target log tables.

The following statements report the status of the load jobs for the staging and target databases:

SELECT JOB_DESC, JOB_NAME, JOB_STATUS FROM CQ10A901.TMP_SBLLOG_STG WHERE JOB_NAME 
LIKE 'LKC%';

SELECT JOB_DESC, JOB_NAME, JOB_STATUS FROM CQ10A901.TMP_SBLLOG_TAR WHERE JOB_NAME 
LIKE 'LKC%';

You can alter the preceding statements to report the status of any jobs by changing the JOB_NAME LIKE statement to another prefix.

This following statement checks for failed unload jobs, but can check for any other job by changing the JOB_NAME LIKE statement to use the appropriate prefix.

SELECT JOB_DESC, JOB_NAME, JOB_STATUS FROM CQ10K034.TMP_SBLLOG_STG WHERE JOB_STATUS != 'COMPLETED SUCCESSFUL' AND JOB_NAME LIKE 'LKB%';