Active Assurance Terminologies

To understand Active Assurance, you should be familiar with the terms defined in Table 1.

Table 1: Terminologies Used in Active Assurance
Term	Definition
Test Agent	Test Agent is a software that is installed on your network device, and it acts as a measurement point in your network. These Test Agents are deployed at strategic locations in your network to evaluate the quality of your network by collecting metric data for preconfigured key performance indicators (KPIs). Test Agents generate, receive, and analyze network traffic, and therefore enable you to continuously view and monitor both real-time and aggregated result metrics.
Test Agent Application	Test Agent Application is a software component that can be installed within your network and delivered through a Docker container to run automated tests. Test Agent Application can be deployed in two ways: On network devices—Test Agent Application is supported in Junos OS Evolved ACX routers that are running locally. You can download the software and install on the supported devices. As a standalone Test Agent Application—The Test Agent Application is delivered as a Docker container image by following standard Docker deployment procedures. You can deploy the application in cloud-based or virtualized infrastructures, without the need for direct installation on physical devices.
Plug-in	A plug-in is an extension to the Test Agent. The Test Agents download the plug-in from Paragon Automation, when required. The plug-in sends the actual traffic and performs the work that is specified in the Tasks and Monitors. Example of plug-in are HTTP, DNS, TCP, UDP, and so on.
Test	A Test is a set of verifications that is performed by one or more Test Agents for a finite amount of time. A Test contains of one or more Steps and Tasks. A Test can consist of several Steps that are executed sequentially. Each Step can consist of one or more Tasks that run concurrently. A Test delivers a binary result—Pass or Fail.
Monitor	A Monitor is a set of verifications that is performed by one or more Test Agents for an infinite amount of time. A Monitor can contain multiple Tasks, which run in parallel and can continuously monitor the KPIs that are defined in the Monitor. A Monitor delivers a time-framed result as it runs continuously.
Task	A Task is a part of Tests and Monitors. When you run a Test or a Monitor, a single task sends a set of Measurements to the Test Agent. The Task also has metrics Evaluation Criteria, which is passed on to the underlying Measurements.
Step	A Step runs a set of parallel Tasks. A Test can have one or more Steps that are executed sequentially, while a monitor has a single Step.
Evaluation Criteria	Configure customized evaluation criteria for metrics by using threshold expressions. On violation of any configured expression, Paragon Automation generates events.
Measurement	When you run a Test or a Monitor, Paragon Automation instructs the Test Agents to send or receive traffic using the selected protocols through measurements. Each measurement produces one or more Streams of metrics.
Stream	Stream represents the stream of data between a Test Agent and an endpoint in a single direction. Depending on the protocol used, an endpoint could be a webserver (HTTP), or another peer Test Agent (TCP). The Tests or Monitors evaluate these Streams and the summary of the Test or Monitor results is displayed on the Test Details (Observability > Active Assurance > Tests >`Test- Name` ) or Monitor Details (Observability > Active Assurance > Monitors > `Monitor-Name`) pages. By looking at these Stream graphs, you can identify what caused the violation.
Errored second	An Errored Second is an aggregation value that is stored as a time-series metric as part of a Stream. It contains information about the number of seconds that the error (threshold violations) has occurred.
Tags	You can configure Tags for a Test or a Monitor in the `key:value` format. A tag is a key-value pair in which the key signifies a category for which you configure a value. The value is an identifier for the category. Examples of key-value pairs are device and device name (edgedevice:acx7000), site and site name (site:bangalore).