Overview of MQ Accounting Data

Introduction to metrics found in MQ Accounting records and ways value can be derived from this typically large data source by viewing the data at various levels of detail including connection type and caller details.

More MQ Accounting Videos

1. Overview of MQ Accounting Data
2. Viewing Accounting Data by Queue Level
3. Viewing Accounting Data by Connection Type
4. Selected Accounting Data Metrics – Part 1
5. Selected Accounting Data Metrics – Part 2
6. Sample MQ Statistics and Accounting Dashboards

Video Transcript

So now let’s go ahead and transition and talk about learning from the MQ Accounting data. So the SMF type 116 records contain detailed task-level data, and most of the accounting data is only generated when Class 3 is specified on the start trace accounting command. The default Class 1 data provides only a very limited set of metrics.

So gaining value from really any large data source is typically achieved through focusing the analysis on a subset of the data. So this is a list of many levels at which the accounting data can be viewed. One common line of analysis is activity for specific queues, including depth of messages on a queue or time that messages spend on a queue. And then another common focus for MQ accounting data, as is also the case with Db2, is by connection type, differentiating characteristics of the activity across various callers of MQ. And the remaining items there under the second column indicate more ways that the accounting data can be subdivided, and typically those will follow an initial focus by connection type.

So the availability of those additional levels of detail differ by connection type. So if the MQ work coming from batch or TSO, then in the connection name field, you get the address space name. The channel initial work can be viewed by the CHIN address space name as well as the channel name and IP address. For CICS, in addition to region name, you’re likely going to want to view the MQ work by transaction ID. Command server work can be viewed by queue manager. And the workload coming from IMS can be viewed by IMS subsystem name and also by program specification block, which often provides an application-oriented view of that work. And then for reference, here are examples of the types of metrics that are present in the accounting data.