DataDog · brett0000FF · Apr 29, 2025 · May 2, 2025 · May 2, 2025 · May 6, 2025
@@ -3865,31 +3865,36 @@ menu:
       identifier: monitor_endpoints
       parent: endpoint_observability
       weight: 803
+    - name: Service Discovery
+      url: /tracing/service_discovery/
+      identifier: service_discovery
+      parent: tracing
+      weight: 9
     - name: Dynamic Instrumentation
       url: dynamic_instrumentation/
       identifier: dyninst
       parent: tracing
-      weight: 9
+      weight: 10
     - name: Enabling
       url: dynamic_instrumentation/enabling
       identifier: dyninst_enable
       parent: dyninst
-      weight: 901
+      weight: 1001
     - name: Expression Language
       url: dynamic_instrumentation/expression-language
       identifier: dyninst_explang
       parent: dyninst
-      weight: 902
+      weight: 1002
     - name: Live Debugger
       url: tracing/live_debugger/
       identifier: live_debugger
       parent: tracing
-      weight: 10
+      weight: 11
     - name: Error Tracking
       url: tracing/error_tracking/
       parent: tracing
       identifier: tracing_error_tracking
-      weight: 11
+      weight: 12
     - name: Error Tracking Explorer
       url: tracing/error_tracking/explorer
       parent: tracing_error_tracking
@@ -3899,102 +3904,102 @@ menu:
       url: tracing/error_tracking/issue_states
       parent: tracing_error_tracking
       identifier: tracing_error_tracking_states
-      weight: 1102
+      weight: 1202
     - name: Error Grouping
       url: tracing/error_tracking/error_grouping
       parent: tracing_error_tracking
       identifier: tracing_error_tracking_error_grouping
-      weight: 1103
+      weight: 1203
     - name: Monitors
       url: tracing/error_tracking/monitors
       parent: tracing_error_tracking
       identifier: tracing_error_tracking_monitors
-      weight: 1105
+      weight: 1205
     - name: Identify Suspect Commits
       url: tracing/error_tracking/suspect_commits
       parent: tracing_error_tracking
       identifier: tracing_error_tracking_suspect_commits
-      weight: 1106
+      weight: 1206
     - name: Exception Replay
       url: tracing/error_tracking/exception_replay
       parent: tracing_error_tracking
       identifier: tracing_error_tracking_exception_replay
-      weight: 1107
+      weight: 1207
     - name: Troubleshooting
       url: error_tracking/troubleshooting
       parent: tracing_error_tracking
       identifier: tracing_error_tracking_troubleshooting
-      weight: 1108
+      weight: 1208
     - name: Data Security
       url: tracing/configure_data_security/
       parent: tracing
       identifier: tracing_data_security
-      weight: 12
+      weight: 13
     - name: Guides
       url: tracing/guide/
       parent: tracing
       identifier: tracing_guides
-      weight: 13
+      weight: 14
     - name: Troubleshooting
       url: tracing/troubleshooting/
       parent: tracing
       identifier: tracing_troubleshooting
-      weight: 14
+      weight: 15
     - name: Tracer Startup Logs
       url: tracing/troubleshooting/tracer_startup_logs
       identifier: tracing_troubleshooting_startup_logs
       parent: tracing_troubleshooting
-      weight: 1401
+      weight: 1501
     - name: Tracer Debug Logs
       url: tracing/troubleshooting/tracer_debug_logs
       identifier: tracing_troubleshooting_debug_logs
       parent: tracing_troubleshooting
-      weight: 1402
+      weight: 1502
     - name: Connection Errors
       url: tracing/troubleshooting/connection_errors
       identifier: tracing_troubleshooting_connection_errors
       parent: tracing_troubleshooting
-      weight: 1403
+      weight: 1503
     - name: Agent Rate Limits
       url: tracing/troubleshooting/agent_rate_limits
       identifier: tracing_troubleshooting_rate_limits
       parent: tracing_troubleshooting
-      weight: 1404
+      weight: 1504
     - name: Agent APM metrics
       url: tracing/troubleshooting/agent_apm_metrics
       identifier: tracing_troubleshooting_apm_metrics
       parent: tracing_troubleshooting
-      weight: 1405
+      weight: 1505
     - name: Agent Resource Usage
       url: tracing/troubleshooting/agent_apm_resource_usage
       identifier: tracing_troubleshooting_agent_usage
       parent: tracing_troubleshooting
-      weight: 1406
+      weight: 1506
     - name: Correlated Logs
       url: tracing/troubleshooting/correlated-logs-not-showing-up-in-the-trace-id-panel
       identifier: tracing_troubleshooting_correlated_logs
       parent: tracing_troubleshooting
-      weight: 1407
+      weight: 1507
 
8000
    - name: PHP 5 Deep Call Stacks
       url: tracing/troubleshooting/php_5_deep_call_stacks
       identifier: tracing_troubleshooting_php_5_deep_call_stacks
       parent: tracing_troubleshooting
-      weight: 1408
+      weight: 1508
     - name: .NET diagnostic tool
       url: tracing/troubleshooting/dotnet_diagnostic_tool
       identifier: tracing_troubleshooting_dotnet_diagnostic_tool
       parent: tracing_troubleshooting
-      weight: 1409
+      weight: 1509
     - name: APM Quantization
       url: tracing/troubleshooting/quantization
       identifier: tracing_troubleshooting_quantization
       parent: tracing_troubleshooting
-      weight: 1410
+      weight: 1510
     - name: Go Compile-Time Instrumentation
       url: /tracing/troubleshooting/go_compile_time
       identifier: tracing_troubleshooting_go_instrumentation
       parent: tracing_troubleshooting
-      weight: 1411
+      weight: 1511
     - name: Continuous Profiler
       url: profiler/
       pre: profiling-1

@@ -0,0 +1,138 @@
+---
+title: Service Discovery
+further_reading:
+- link: "/agent/fleet_automation/"
+  tag: "Documentation"
+  text: "Fleet Automation"
+- link: "/tracing/trace_collection/automatic_instrumentation/single-step-apm/"
+  tag: "Documentation"
+  text: "Single Step APM Instrumentation"
+- link: "/data_streams/"
+  tag: "Documentation"
+  text: "Data Streams Monitoring"
+---
+
+## Overview
+
+Service Discovery provides visibility into the monitoring coverage of your services within Datadog. It automatically discovers services running across your infrastructure, helps you identify potential observability gaps, and provides relevant information to triage and take action.
+
+Service Discovery is accessed through the **[Fleet Automation > Services][1]** page.
+
+{{< img src="/tracing/service_discovery/service_discovery.png" alt="Datadog Fleet Automation Services page showing a list of discovered services not sending traces. Each service entry includes tags, container count, CPU and memory usage, and network activity." style="width:100%;" >}}
+
+### Key benefits
+
+- **Discover all services**: View both monitored and unmonitored services running in your fleet in one centralized location.
+- **Demystify services**: Understand the potential importance of unmonitored services using contextual information like infrastructure footprint, resource consumption, network activity, configuration details, and tags.
+- **Close observability gaps**: Receive recommendations and guided instructions to instrument services with APM or enable other relevant Datadog products such as [Data Streams Monitoring (DSM)][3].
+- **Triage effectively**: Prioritize which services to monitor using sortable metadata columns, facet filtering, and an organization-wide ignore list for noisy or irrelevant services.
+
+### How it works
+
+1. The Datadog Agent inspects running processes and container metadata on supported hosts.
+2. Processes are automatically grouped into services based on a defined naming hierarchy (see [Discovered service naming](#discovered-service-naming).
+3. Datadog checks if discovered services are emitting traces to Datadog APM.
+4. Services are displayed in **Fleet Automation > Services** under either **Monitored with APM** or **Not Sending Traces**.
+5. Contextual metadata (CPU, memory, network I/O, tags, infrastructure links) is associated with each service to aid prioritization.
+6. Short-lived processes (running for less than 1 minute) are automatically ignored to reduce noise.
+
+### Discovered service naming
+
+Service Discovery automatically identifies and names services based on a priority order of available identifiers:
+
+1. Existing Datadog service name set using [Unified Service Tagging][5].
+2. Container labels (for containerized services).
+3. Language-specific manifest files (for example, `package.json` for Node.js).
+4. Command-line arguments and process information.
+
+
+For Java enterprise web applications (specifically JBoss, Websphere, Tomcat, Jetty, and Weblogic), multiple web applications running in the same process are displayed as individual services in the unmonitored services list, with visual indicators showing they belong to the same process.
+
+## Requirements
+
+Service Discovery requires the Datadog Agent and is supported on specific operating environments:
+
+| Environment                   | Minimum Agent Version |
+|-------------------------------|-----------------------|
+| Linux Hosts (x86-64)          | `[7.xx.x+]`           |
+| Docker Containers (on Linux)  | `[7.xx.x+]`           |
+| Kubernetes (Helm Chart)       | `[7.xx.x+]`           |
+| Kubernetes (Datadog Operator) | `[7.xx.x+]`           |
+
+<div class="alert alert-info">Service Discovery supports Linux (x86-64) environments only. Windows, ARM architectures, and serverless environments are not supported.</div>
+
+## Setup
+
+Enable Service Discovery by installing the latest version of the Datadog Agent with the Service Discovery feature turned on.
+
+To enable Service Discovery on Linux hosts, install the latest version of the Datadog Agent and use the toggle in the installation UI:
+
+1. Navigate to [**Fleet Automation > Install Agents**][6].
+2. Select your platform.
+3. Follow the installation instructions, and ensure that the **Service Discovery** toggle is turned on.
+
+Allow a few minutes for data to appear on the **Fleet Automation > Services** page.
+
+## Explore discovered services
+
+Navigate to **[Fleet Automation > Services][1]** in Datadog to view discovered services. The page presents two main views:
+
+- **Not Sending Traces**: Services discovered by the Agent that are not sending trace data to Datadog APM.
+- **Monitored with APM**: Services actively sending trace data to Datadog APM.
+
+### Monitored with APM
+
+This view lists services already instrumented and sending trace data to Datadog APM. It provides insights into the current monitoring state:
+
+- **Type**: Shows the service type with its corresponding icon.
+- **Service**: Displays the service name.
+- **APM SDK**: Shows the versions of Datadog tracing libraries used by the service instances.
+- **Telemetry**: Indicates the types of telemetry being collected (Traces, Logs, USM, DSM, Profiling).
+
+### Not sending traces
+
+This view lists services detected on your infrastructure that are not monitored by Datadog APM. Understanding these services is the first step toward closing observability gaps.
+
+For each unmonitored service, Datadog provides contextual information to help you assess its importance and prioritize instrumentation:
+
+- **Service**: The name assigned to the service, based on the naming hierarchy described above.
+- **Infra**: An overview of the hosts or containers the service is running on, displayed as container icons with counts. Clicking this provides links to the relevant infrastructure components in Datadog.
+- **CPU Usage**: Shows CPU cores usage.
+- **Memory Usage**: Shows memory usage.
+- **Bytes Received**: Shows incoming network traffic.
+- **Bytes Sent**: Shows outgoing network traffic.
+
+#### Triaging unmonitored services
+
+- **Sorting**: By default, services are sorted by infrastructure footprint (descending). You can sort by other columns like CPU, Memory, or Network Activity to prioritize based on resource usage or traffic.
+- **Filtering**: Use the search bar and facets to narrow down the list.
+- **Ignoring Services**: If a discovered service is noisy or not relevant for monitoring (such as temporary utilities or test workloads), hover over the service and click **Ignore**. Ignored services are hidden from the main list but can be viewed and restored from the **Ignored Services** toggle.
+
+**Enabling APM:**
+
+For services you decide to monitor, click the **Enable APM** button. Datadog provides instructions to enable APM using **Single Step Instrumentation** for supported languages.
+
+## Troubleshooting
+
+- **No services listed under "Not Sending Traces"**:
+  - Verify the Datadog Agent version meets the minimum requirement.
+  - Confirm that Service Discovery is correctly enabled during Agent installation.
+  - Ensure services have been running for more than 1 minute on supported platforms.
+  - Allow a few minutes after Agent installation for data to populate.
+- **Some expected services are missing**:
+  - Check if the service runs for less than 1 minute.
+  - Confirm the service is running on a supported platform (Linux x86-64).
+
+If issues persist, collect an [Agent flare][4] and contact [Datadog Support][2].
+
+## Further reading
+
+{{< partial name="whats-next/whats-next.html" >}}
+
+[1]: https://app.datadoghq.com/fleet/services
+[2]: /help/
+[3]: /data_streams/
+[4]: /agent/troubleshooting/send_a_flare/
+[5]: /getting_started/tagging/unified_service_tagging/
+[6]: https://app.datadoghq.com/fleet/install-agent/latest?platform=overview
+
@@ -53,10 +53,7 @@ Capture observability data from in-house code or complex functions that aren't c
 
 To learn more, see [custom instrumentation][6].
 
-{{< callout url="https://www.datadoghq.com/product-preview/service-discovery/" btn_hidden="false" header="Service discovery is in Preview">}}
-Service discovery provides complete visibility into the current state of application monitoring, highlighting any major gaps or broken traces in your system. 
-{{< /callout >}}
-
+<div class="alert alert-info"><a href="/tracing/service_discovery/">Service Discovery</a> provides visibility into the monitoring coverage of your services within Datadog. It automatically discovers services running across your infrastructure, helps you identify potential observability gaps, and provides relevant information to triage and take action.</div>
 
 ## APM setup tutorials