Link Search Menu Expand Document Documentation Menu
Documentation

Observability workspace apps

Introduced 3.5

In an Observability workspace, OpenSearch Dashboards provides apps for exploring logs, traces, and metrics in Discover, monitoring application performance, tracking service-level objectives (SLOs), and monitoring generative AI agents. Each app is enabled by a setting in opensearch_dashboards.yml, which is listed on the app’s page.

Prerequisites

To enable all Observability workspace apps, add the following settings to opensearch_dashboards.yml, and then restart OpenSearch Dashboards:

workspace.enabled: true
explore.enabled: true                  # Logs
explore.discoverTraces.enabled: true   # Traces, APM, and SLOs
explore.discoverMetrics.enabled: true  # Metrics and APM
explore.agentTraces.enabled: true      # Agent monitoring
datasetManagement.enabled: true        # Datasets and Correlations
observability.slo.enabled: true        # SLOs

To enable only some of the apps, see the Prerequisites section of each app’s page for the settings that the app requires.

Workspaces have additional requirements, such as disabling multi-tenancy. For more information, see Enabling workspaces.

Creating an Observability workspace

To create an Observability workspace, follow these steps:

  1. On the OpenSearch Dashboards home page, in My workspaces, select Create workspace on the Observability card.
  2. In Workspace details, enter a Name and, optionally, a Description.
  3. In Use case and features, confirm that Observability is selected.
  4. If you use multiple data sources, select the data sources to associate with the workspace.
  5. Select Create workspace.

For more information, see Create a workspace.

Exploring and monitoring your data

After you create the workspace, use the apps as follows:

  1. Connect your data:
    • Create datasets for your log and trace indexes. The Logs and Traces pages, Correlations, and APM use datasets to find your data.
    • To use Metrics, APM, or SLOs, connect a Prometheus data source.
  2. Explore your data:
    • Use Logs to search, filter, and visualize log data using Piped Processing Language (PPL).
    • Use Traces to search and filter spans and view a single trace as a timeline of its spans.
    • Use Metrics to query and visualize Prometheus metrics using PromQL.
    • Use Correlations to link a trace dataset to logs datasets so that you can view the log entries related to a trace.
  3. Monitor your applications:
    • Use application performance monitoring (APM) to view rate, error, and duration (RED) metrics for each service and a topology map of service dependencies.
    • Use SLOs to define availability and latency targets for your services and track their error budgets.
  4. Monitor your AI agents:
    • Use Agent monitoring to view the LLM calls, tool invocations, and token usage of generative AI agents.
  5. Share your results:
    • Add visualizations from the Logs page to dashboards.
    • Use Notebooks to combine Markdown text, SQL and PPL queries, and visualizations into reports and runbooks.
350 characters left

Have a question? .

Want to contribute? or .