If you run low-level APPC traces (see the problem determination guide for your specific APPC product for information on how to run traces), you can get access to detailed sense data that can help you determine the cause of a particular problem.
Sense data is an SNA-defined encoding of error information. In LU 6.2 products, sense data can be found as part of internal SNA flows, error logs, error messages, and traces. In IBM's OS/2 Communications Manager, for example, it is even returned in the verb control block of an Allocate call (when using the native APPC API) when it fails because of an ALLOCATION_FAILURE condition. The sense data is a 4-byte field (as shown in the following figure) that includes:
<--- Byte 0 ---><--- Byte 1 ---><--- Byte 2 ---><--- Byte 3 --->
<---Category---><---Modifier---><----Sense-code specific------->
information
<----------Sense Code---------->
<------------------Sense Data---------------------------------->
Together, the category byte 0, the modifier byte 1, and the sense-code specific bytes 2 and 3 hold the sense data defined for the exception condition that has occurred. The sense codes are discussed on the following pages. For information about sense data not documented here, refer to SNA Formats, IBM document number GA27-3136, or SNA Network Product Formats, IBM document number LY43-0081.
| Note: | This information can also be found on the diskette accompanying the book CPI-C Programming in C: An Application Developer's Guide to APPC, published by McGraw-Hill, New York, NY, ISBN 0-07-911733-3. |