Before you begin:

You need to document the problem that you have and how it presents itself

Perform these steps to use the tables:

First, determine which type of symptom best describes the problem that you are troubleshooting and locate that type in the first column, Symptom Type, in Table 1.
_________________________________________________________________
If you find a matching symptom type:
1. Go to the topic listed in the second column, Refer To Topic..., in Table 1. This is the appropriate troubleshooting table.
2. In the appropriate table, locate the specific RACF® symptom for the problem (for example, Abends). In the second column, Recommended Diagnostic Procedures, in Table 1, follow the steps for diagnosing the problem.
_________________________________________________________________
If the problem has several symptoms, you should look up each symptom in the table. For example, some problems cause an abend and an I/O error message.
_________________________________________________________________

Table 1. Index to Troubleshooting Tables for RACF
Symptom Type	Refer To Topic ...
Abend	Abend troubleshooting table
Documentation error	Documentation troubleshooting table
Incorrect output	Incorrect output troubleshooting table
Logon problems	Logon problems troubleshooting table
Loop	Loops troubleshooting table
Messages	Messages troubleshooting table
Performance degradation	Performance problems troubleshooting table
Return code from RACF	RACF return codes troubleshooting table
Wait or hang	Waits troubleshooting table

If none of the symptoms in the list matches your problem, try these diagnostic procedures:

Check whether any error messages were issued for the user request and make sure that message IDs are included with the messages.
If a batch job produced the output, check that the job statement had MSGLEVEL=(1,1) specified. If a TSO/E user had the problem, check that the user profile had these options: PROFILE WTPMSG MSGID

_________________________________________________________________
If you find an error message that might relate to the problem, see Table 1.
_________________________________________________________________
If you do not have a system dump for the problem, and you want to request one, see Obtaining a system dump.
_________________________________________________________________
If you have a system dump for the problem, display or print the symptom record in the dump.
Use the IPCS VERBEXIT LOGDATA and VERBEXIT SYMPTOMS subcommands.

Note: SYMPTOMS has some prerequisite VERBEXITs, such as DAEDATA, which provide the abend and non-abend symptoms for the dump. You can also use the BLSCSCAN CLIST with IPCS to obtain the prerequisites. For details, see z/OS MVS IPCS User's Guide.

If the symptom record indicates an abend, loop, or wait, use the corresponding entry in the list of symptom types in Table 1.

_________________________________________________________________
If you have SYS1.LOGREC output for the problem, or a dump with LOGDATA output, check the LOGREC records related to the problem. If you have a software LOGREC record for an abend, use Table 1.
_________________________________________________________________

You know you are done when you have matched a system type with the type of problem you are having.

If you still do not have a symptom type,

Check the general problem analysis advice in z/OS MVS Diagnosis: Reference.
Read the information in Making sure this is a RACF problem.
If you need additional assistance, report the problem to IBM®.