|
By specifying the instruction TRACE RESOURCES ON, NAPA records data for all resources. Using a TYPES parameter, you can selectively capture data about a specific resource type. For example, TRACE RESOURCES ON TYPES (N2MEMORY, CPUTIME) captures the CPU resources and the total number of bytes of memory used by NOMAD in a particular activity.
Record Types
NAPA records both snapshot and interval values. A snapshot value provides information about the current state, such as current memory usage. Interval values measure the amount of resource used over a selected section of the application, such as CPU time used since beginning a load procedure. The selected sections to be measured are simply marked by inserting the commands, TRACE RESOURCES ON and TRACE RESOURCES OFF, before and after the section of code to be monitored.
Flexible Monitoring Commands
NAPA features flexibility in monitoring applications. Analysis points can be indicated to capture recorded values at significant points in the application. For example, a snapshot of memory usage values can be captured by inserting TRACE RESOURCES NOW before and after a PURGE ALL or SCREEN DELETE to measure the effectiveness of these operations.
TRACE RESOURCES TEXT enables specific analysis points to be documented to annotate the output file. This facility aids in the analysis of the monitored data.
In addition to specifying the resource type to be monitored, the processes for which analysis occurs can be indicated. Points such as procedural commands, conversational commands and call commands can be tracked individually or in combination.
Information generated by NAPA can be stored in a database and compared against results gathered periodically, making it easy to monitor application performance over time. Monitoring commands embedded in an application can be turned on and off by using the commands, TRACE RESOURCES, ENABLE and DISABLE. By enabling or reactivating the embedded commands whenever recorded output is desired, an application can be monitored periodically without modifying the application code.
Captured Data
The data produced by NAPA can be captured in several ways. When an analysis point occurs, a record of monitored data is written to an output file or dataset. Multiple Schemas are provided in the documentation that give examples of how the recorded data can be structured to allow NOMAD's use for further reporting and analysis.
For data that would best be presented graphically, NOMAD makes it easy using CREATE FOR SAS and ICU. In addition, values from an analysis point can alternately be set into NOMAD &variables, or displayed at a workstation within a NOMAD window.
Monitoring Shared Database IDs
NAPA also provides a special facility for tracking the resources used by shared database servers (SHARE TRACE ownerid). When NAPA is used for a shared database ID, an analysis point occurs when the processing of each transaction for each driver is complete. The resource monitoring can be specified in the control statements of an application, or can be interactively initiated and controlled by the Database Administrator.
Monitoring User-Defined Resources
Additional information, such as total number of users on a system and other environmental considerations not automatically captured by NAPA, can be tracked using the NOMAD Generalized Exit Method (GEM). An installation-specific, user-provided exit program can be called by GEM whenever an analysis point is desired. The GEM user exit traces multiple values of either the snapshot or interval type.
On-site Consultation
As an added benefit, two days of on-site Select consulting are included to help you get a fast start using NAPA in your organization. Through a combination of lectures and demonstrations, Select consultants explain NAPA's capabilities as a diagnostic tool, and show how to use NAPA to measure portions of a sample application. You'll also be guided in using NAPA to measure an application of your choice.
|
