Comparison of READA and READS
The READS service collects trace data at the time you request it.
The READS service collects trace data at the time you request it. Such immediate collection is termed synchronous. In contrast, the READA service does not collect any data. It simply copies into your exec the trace records that have accumulated in a Db2 trace buffer.
READA differs from READS in the following respects:
- To use READA, your REXX exec must start a Db2 trace to an OPn destination and must leave it active. The -START TRACE command obtains ownership of an OP buffer and starts collection of the trace events (IFCIDs) you specify on the -START TRACE command. Db2 writes trace records to these buffers. Your monitor exec obtains this data later (asynchronously) when you issue a READA request.
- There is no specification of IFCIDs on the READA request. Instead, you specify, through various operands of the -START TRACE command, what IFCIDs you want to collect. You can restrict the trace events that are recorded to IFICIDs within classes or to particular IFCIDs.
- You do not qualify what threads to monitor on the READA request. (The set of qualifying variables whose names start with the letters WQAL pertain only to READS requests). You may however limit tracing to particular plans and authorization IDs through the constraint block operands of the START TRACE command.