Delay in displaying OpenTelemetry metric dimensions in Dynatrace

ondrejdubaj — Mon, 17 Feb 2025 08:41:18 GMT

Problem

When adding new dimensions to a metric, the newly added dimensions may not immediately appear in the Dynatrace dropdown menu, even if the data points are present.

Explanation

The delay in displaying newly added dimensions in the Dynatrace dropdown menu is primarily due to the propagation time required for the new data to be fully integrated into the system.

When ingesting OpenTelemetry metrics, both the metric’s data points and its metadata are ingested together.

This means that when metrics are ingested successfully (with no errors returned from the Dynatrace API), you can be sure that your data is stored securely. There is no possibility that metrics are ingested without their metadata.

What can happen is that it may take some time until the metric’s metadata is fully propagated and available for use by the suggestions/ auto-completion mechanisms in the Dynatrace web UI.

The fact that the dimensions are not showing does not mean the data is lost. You can still manually use them, even though they are not yet being shown.

Impact

The delay in the dropdown menu does not affect the functionality of the metrics. The data is fully ingested and operational. You can still perform operations on the metric using the new dimension by manually typing the dimension name in the advanced mode - this confirms that the data is present and functional even if it is not yet visible in the dropdown.

Resolution

Use the advanced mode to manually specify the new dimension name and continue their operations without waiting for the dropdown menu to update.
Use DQL to manually add the name of the dimension they would like to split (e.g. in Notebooks)
Specify the metadata manually, see custom metric metadata.

What's Next?

If manually querying the dimension as instructed above does not work or dimensions don’t show up after some time, make sure you are not running into the limits and also look at your logs for API errors or warnings.