2. COST ACCOUNTING CONCEPTS FOR IS ORGANIZATIONS › 2.3 Resource Accounting › 2.3.4 Transaction Residency Time

2.3.4 Transaction Residency Time

Transaction residency time attempts to measure the time a
transaction is active in the processor for interactive
systems such as CICS and IMS.  The diagram below helps
explain this measurement:



                    task         task
       _____      attached     detached        ______
      |     |         |            |          |      |
      |_____|         |            |          |______|
         |         |  |            | |
 input   |            |            |              |  output
 message  -------->|  |<__________>| | ------------ message
                        Transaction
                         Residency



When a message enters the system, it is transmitted to the
CPU, where it resides until the processing completes.  The
period between the time of input and the time the task is
attached is not included in transaction residency time, nor
is the period between the time the task is detached from the
CPU and the time the output message is received.

Because transaction residency time is dependent on the
multiprogramming mix, it varies with the performance of the
computer and the number of other transactions in the system.
Since the same task does not always have the same transaction
residency time, it is an inconsistent, and therefore
unsatisfactory, method of measurement.