Performance Analytics for Managed Service Providers

[Robert Ninness]

Performance Analytics on any instance can be simple: taking out of the box content packs, or complex: custom KPIs and dashboards. With the potential for multiple process models on a domain separated instance, additional planning and setup may be required.

In the attached paper I discuss and provide guidance to implementers looking to dive into Performance Analytics on a domain separated instance.

Please download and read the attached document to get started with Performance Management for Managed Service Providers.

Some excerpts:

The ideal solution is to stay out of the box as much as possible. Performance Analytics content packs, distributed along-side applications, are designed cooperatively to work with no additional configuration. Changes made to the process model of an application, may also need to be reflected in the configuration of the Performance Analytics content pack. . . .

Performance Analytics uses the concepts of data and process separation to simplify configuration and ensure that data visibility rules are adhered to when collecting, storing and retrieving scores. The majority of Performance Analytics tables (prefixed with pa_) are Process Separated. Classifying Performance Analytics records as process, as opposed to data, allows domains to inherit a single set of configurations, making setup and ongoing maintenance much simpler. . . .

The data collector can be configured to collect scores across many domains at the same time, using one set of indicator configurations. Data separation on source tables is still respected when retrieving and calculating scores. Just like the source tables, scores are data separated. When a dashboard is loaded, although that dashboard may be available to users across multiple domains, the widgets will only display scores from the user’s domain. . . .