Collecting diagnostic data

If you have a problem that you are unable to solve using the information in this guide, gather the following information that relates to the problem and contact IBM® Software Support for further assistance.

Description of the operation scenario that led to the problem.
Operating system version and release level.
Version, release and fix level of the following members of the monitoring environment:
- IBM Tivoli® Monitoring Server
- Tivoli Enterprise Portal Server
- Monitoring agent
If the Windows operating system crashes, collect the drwtsn32.log and user.dmp files, if available. The files are located in the following path:
\Documents and Settings\All Users\Documents\DrWatson

See the chapter on diagnostic tools in the IBM Tivoli Monitoring Troubleshooting Guide for more information about trace tools.
- Enter the following command at the command prompt to enable the Dr. Watson™ tool as default debugger: drwtsn32 –i 2
- Enter the following command at the command prompt to open the configuration window of the Dr. Watson tool: drwtsn32
Include such other documentation as is recommended in Table 6.
If possible, try to classify your problem according to a component area listed in Table 7 and collect documentation as listed there.

Table 1. Recommended documentation
	Abend	Incorrect output	Performance	Function	Connectivity
LOG	X	X	X	X	X
Dump	X	X	X
Screen shots		X		X
Traces			X
cinfo output				X
TEPS database export				X
LISTDATA output		X		X
SMF record dump		X
BVIR dump	X	X
RKS3DATA dump	X	X

Table 2. Requirement matrix based on problem type and product component
Component area	Abend	Incorrect output	Performance	Function loss	Connectivity
Application monitoring	RKLVLOG Dump	RKLVLOG Dump Screen Shots¹	N/A	RKLVLOG Screen Shots cinfo output TEPS Db Export	N/A
Cache CU stats	RKLVLOG Dump	RKLVLOG Dump Screen Shots LISTDATA⁶	RKLVLOG Dump Traces²	RKLVLOG Screen Shots cinfo output LISTDATA⁶	N/A
Tape & VTS	RKLVLOG Dump SMF Record Dump⁴ BVIR Dump⁴	RKLVLOG Dump Screen Shots SMF Record Dump⁴ BVIR Dump⁴	N/A	RKLVLOG Screen Shots cinfo output	N/A
DASD performance	RKLVLOG Dump	RKLVLOG Dump Screen Shots	N/A	RKLVLOG Screen Shots cinfo output	N/A
DASD Space	RKLVLOG Dump	RKLVLOG Dump Screen Shots	RKLVLOG Dump Traces²	RKLVLOG Screen Shots TEPS Db Export	N/A
DADb	RKLVLOG Dump³	RKLVLOG Dump³ Screen Shots¹	RKLVLOG Dump³ Traces²	RKLVLOG Screen Shots TEPS Db Export	N/A
Dataset group	RKLVLOG Dump³	RKLVLOG Dump³ Screen Shots¹	RKLVLOG Dump³ Traces²	RKLVLOG Screen Shots cinfo output TEPS Db Export	N/A
SMS	RKLVLOG Dump	RKLVLOG Dump Screen Shots	N/A	RKLVLOG Screen Shots	N/A
HSM	RKLVLOG Dump	RKLVLOG Dump Screen Shots	N/A	RKLVLOG Screen Shots	N/A
RMM	RKLVLOG Dump³	RKLVLOG Dump³ Screen Shots¹	RKLVLOG Dump³ Traces²	RKLVLOG Screen Shots cinfo output TEPS Db Export	N/A
Toolkit	RKLVLOG Dump RKS3DATA Dump	RKLVLOG Dump Screen Shots¹ RKS3DATA Dump	N/A	RKLVLOG Screen Shots cinfo output TEPS Db Export	N/A
OM II Interface	RKLVLOG Dump	RKLVLOG⁵ Dump Screen Shots	RKLVLOG⁵ Dump³ Traces²	RKLVLOG Screen Shots	RKLVLOG⁵
Historical	RKLVLOG Dump	RKLVLOG Dump Screen Shots¹	RKLVLOG Dump³ Traces²	RKLVLOG Screen Shots	N/A

Note:

Should include screen shots of relevant definition dialogs, as well as the errant workspaces. For instance, documentation for incorrect output in the Dataset Group Summary should contain screen shots of the dialog where the groups are defined, as well as the workspace showing the incorrect values.
See the section on Traces for information about setting traces for specific component areas.
Some components store data in data spaces and the dump created for the abend may not contain the data space, therefore you may need to create a separate console dump to capture the data space and send it together with the abend dump. See the section on Dump Creation for more information.
If your problem involves an IBM VTS that pre-dates the TS7700 series, statistics are extracted from SMF Type 94 records. Please use IFASMFDP to dump all Type 94 records from one of your MAN datasets and submit it with your other documentation. See the section on Dumping SMF Records for more information.
If your problem involves an IBM TS7700 series VTS, statistics are extracted by a BVIR dataset which is created every interval. Please use a batch job to mimic this collection process and submit the output BVIR data set to support along with your other documentation. See the section on Creating a BVIR data set in batch mode for more information.
OMEGAMON for Storage gathers cache information using IBM IDCSS01 service. A good way to determine whether a problem in collecting cache statistics is due to an OMEGAMON bug is to use another utility that relies on IDCSS01. IDCAMS LISTDATA command does this. See the section on Running LISTDATA for information on setting up a job to test statistics collection on a controller.