z/OS Communications Server: ACF/TAP Trace Analysis Handbook
|
Previous topic |
Next topic |
Contents |
Contact z/OS |
Library |
PDF
Scanner interface trace z/OS Communications Server: ACF/TAP Trace Analysis Handbook GC27-3645-00 |
|
|
If your lines are attached to a 3746 Model 900, see CSS adapter trace. The access method, the scanner microcode, and NCP share the SIT. Although the access method controls the trace, the scanner microcode gathers information in the trace records and passes it to NCP. NCP then passes it to the host access method. Scanner interface trace (SIT) records the operating parameters of a line whenever the communication controller sends or receives data. SIT records outbound data after NCP processes it. It also records inbound data before it is passed to NCP. If you run SIT on an inactive line attached to the communication controller, this SIT does not collect data. See NCP and EP Reference, LY43-0029 for information about how SIT operates. SIT or line traces can be active for up to eight half-duplex (HDX) lines or four full duplex (FDX) lines at a time for each communication controller. For high-speed links that use a type 2, type 3, or type 4B LIC, you can activate only two SIT or line traces for HDX lines on a scanner. For a high-speed link that is an FDX line and uses a type 2, type 3, or type 4B LIC, you can activate a SIT on only one scanner on an IBM® 3745 at a time (one line trace for each communication controller). You can trace only 40 bytes of data for each IBM 3745 using HPTSS lines.
The number of allowable active SITs is specified during NCP generation. In a multisystem network, where the communication controller is connected to more than one host processor, the number of active traces allowed is distributed among the connected host processors. As the number of active SITs increases, system performance becomes less efficient.
|
Copyright IBM Corporation 1990, 2014 |