Overview

A pipeline consists of components that collect, process, and route your observability data. You can track the status of your pipelines and components in the following ways:

• View health graphs of your pipelines, Workers, and components (sources, processors, and destinations).
• Enable out-of-the-box monitors that alert you if:
  • An Observability Pipelines Worker has high CPU or memory usage, or is dropping data.
  • A component is emitting errors.
  • A defined quota has been reached.
• Create your own dashboards, notebooks, and monitors with the available Observability Pipelines metrics.

View the status of your pipelines

1. Navigate to Observability Pipelines to see how many bytes your pipelines are receiving and sending out.
2. Select a pipeline.
3. Click the Health tab to see details about the pipeline and its components. You can view graphs of:
   • How much each component is being used, and the total number of events that the component receives and sends out.
   • The number of requests made to destinations, and the number of errors encountered by those requests.
   • How many events are intentionally and unintentionally discarded.
   • Any changes in the number of requests and errors for each component over the previous week.

You can export a health graph to a dashboard, notebook, or monitor. The exported graph shows you that the metric is grouped by the specific pipeline and component tags.

View the status of your Workers

To view graphs of resource usage and data sent through Observability Pipelines Workers:

1. Navigate to Observability Pipelines.
2. Select a pipeline.
3. Click the Workers tab to see the Workers’ memory and CPU utilization, traffic stats, and any errors.

View the status of your pipeline components

To view metrics for a source, process, or destination:

1. Navigate to Observability Pipelines.
2. Select a pipeline.
3. Click the cog next to the source’s, processor’s, or destination’s name, then select View details. Datadog displays health graphs for the component you selected.
4. If you want to export a graph to an incident, dashboard, or notebook, click the export icon on the graph. The exported graph shows that the metric is grouped by the specific pipeline and component tags.

Out-of-the-box monitors

To see available out-of-the-box monitors:

1. Navigate to Observability Pipelines.
2. Click Enable monitors in the Monitors column for your pipeline.
3. Click Start to set up a monitor for one of the suggested use cases.
   The new metric monitor page is configured based on the use case you selected. You can update the configuration to further customize it. See the Metric monitor documentation for more information.

Further reading

Documentation, liens et articles supplémentaires utiles:

Set up a pipelineDOCUMENTATION Configure a metric monitorDOCUMENTATION Observability Pipelines metricsDOCUMENTATION