opensearch-docs-cn/_observability-plugin/app-analytics.md

layout	title	nav_order
default	Application analytics	80

Application analytics

Application analytics can now create custom Observability Applications to view the availability status of all their systems, where log events can be combined with trace and metric data, into a single view of system health empowering developers and IT Ops to resolve issues faster and with fewer escalations.Application analytics in Observability is where you can select logs, traces, and metrics to be part of an Observability Application that can be monitored for overall health and visualized on one page. This allows you to quickly pivot between logs, traces, and metrics to dig into the source of any issues. Many teams use similar logs and similar traces and need a structured view.

Get started with application analytics

To get started, choose Observability in OpenSearch Dashboards, and then choose Application analytics.

Create an application

1. Choose Create application
2. Enter a name for your application and optionally add a description
3. Do at least one of the following

• Use PPL to specify the base query.

You can't change the base query after the application is created. {: .note }

• Select services & entities from the dropdown or the service map.
• Select trace groups from the dropdown or the table.

1. Choose Create.

Create a visualization

1. Choose the Log Events tab
2. Use PPL to build upon your base query
3. Move to the Visualizations tab to see your visualizations
4. Expand the Save dropdown menu, enter a name for your visualization, then choose Save.

To see your visualizations, move to the Panel tab.

Configure availability

Availability is the status of your application determined by availability levels set on a time series metric.

To create an availability level you must configure the following

• color: The color of the availability badge on the home page.
• name: The text in the availability badge on the home page.
• expression: Comparison operator to determine the availability.
• value: Value to use when calculating availability.

By default, Application analytics shows results from the last 24 hours of your data. To see data from a different timeframe, use the date and time selector.

Time series metric

A time series metric is any visualization that has a query that spans over a timestamp and is a line chart. Users can use the power of the PPL language to define arbitrary conditions on their logs to create a visualization over time.

Example

source = <index_name> | ... | ... | stats ... by span(<timestamp_field>, 1h)

Users can then choose Line in visualization configurations to create a time series metric.