Vector logging

Vector logging is a diagnostic tool that allows you to record the flow of data between CICS® applications and a virtual terminal. You can use it in a development environment to assist in the deployment and running of a service flow that includes Link3270 server adapters.

Stabilized feature: Service Flow Runtime and Service Flow Modeler capability in IBM Developer for z/OS 14.2.3 are stabilized. Consider exposing and orchestrating applications as API services by using z/OS Connect Enterprise Edition, CICS web services, or by writing web applications in Java or Node.js. See also Stabilization notices.

You can set vector logging as an option when a flow is modeled in service flow project tools or you can set it with the flow management transaction CMAN. When the service flow is deployed, the service flow properties file indicates that vector logging will be performed for that specific Link3270 server adapter. Alternatively, if you want to turn vector logging on for a Link3270 server adapter that is already deployed, use the CMAN transaction.

You can select two levels of granularity for vector logging:

Full vector logging: This level of logging records the header structures and the inbound and outbound vectors including any vector data. The vector data can be an inbound or outbound application data structure (ADS), text, or 3270 datastream.
Vector logging trace: This level of logging records the header structures and the inbound and outbound vectors. It does not record any vector data.

When you switch on vector logging, the Link3270 server adapter writes to the active vector log file. Two files, DFHMALVA and DFHMALVB, store vector logging information. One of these files is always active and the other is either empty or contains old data. The Link3270 server adapter starts by writing vector logging data to DFHMALVA. When the file is full, the Link3270 server adapter swaps to continue writing information to DFHMALVB. When DFHMALVB is full, the server adapter moves back to writing to DFHMALVA, deleting the old content.

By using two files, you can analyze the complete vector logging data for a server adapter. A server adapter is not likely to write enough data to fill both files, but if it does happen, the most recent data is always available.

You can access the contents of the vector log file by dumping it using the provided sample JCL job DFHMAMVD. This job runs the dump utility DFHMAVUP, which formats the file so that you can read it to help with problem determination. The dump utility displays the records in chronological order.