Skip to main content
Version: v2.x

Datadog Integration on Hasura Cloud

Introduction

You can export metrics, operation logs and traces of your Hasura Cloud project to your organization's Datadog dashboard. This can be configured on the integrations tab on the project's setting page.

Note

For Hasura Cloud projects, the Datadog Integration is only available on the Standard (pay-as-you-go) tier and above.

More metrics on OpenTelemetry exporter

Try the new OpenTelemetry exporter to get advanced metrics and traces to connect with Datadog. This integration will export the metrics as detailed here.

Configure Datadog integration

Navigate to the integrations tab on project settings page to find Datadog integration.

Configure Datadog Integration

Select the Datadog API region and enter the Datadog API key (can be retrieved by navigating to Datadog's settings page by clicking the Get API Key link), host, service name and tags to associate with exported logs, metrics and traces.

FieldDescription
RegionHasura supports all available Datadog regions, namely US1, US3, US5, EU, US1FED.
API KeyAPI keys are unique to your organization. An API key is required by the Datadog agent to submit metrics and events to Datadog. Click on Read more above the API Key field in the Datadog integration form to get the API key.
HostThe name of the originating host of the log and metrics.
TagsTags associated with your logs and metrics. Default tags project_id and project_name are exported with all logs and metrics. A source tag hasura-cloud-metrics is added to all exported logs.
Service NameThe name of the application or service generating the log events.
Configure Datadog Integration

After adding appropriate values, click Save.

Checking the status of the integration

The green/red dot signifies the status of the integration. Green signifies successful exporting of logs to datadog. When logs are successfully exported, Last Exported is continuously updated, indicating the timestamp of the last log line successfully exported to your Datadog dashboard.

Datadog Integration successfully configured

In case there is an error while exporting logs to datadog, the dot is red and the HTTP status code of the error is displayed right below it.

Datadog Integration successfully configured

View logs

The logs can be viewed in your Datadog dashboard, under the Logs tab. To navigate to the same, click View Logs.

Datadog Integration successfully configured
Logs successfully exported to Datadog

To view only logs exported by Hasura Cloud, filter your logs using host and/or tags you configured with this integration.

Note

Datadog allows ingestion of logs with maximum size 256kB for a single log. If a log exceeds this limit, Datadog will truncate the log at 256kB.

View metrics

The integration exports the following five metrics to your Datadog dashboard:

Metric ExportedMetric Name in Datadog
Average number of requestsaverage_requests_per_minute
Average request execution timeaverage_execution_time
Success rate of requestssuccess_rate
Active subscriptionsactive_subscriptions
Number of websockets openwebsockets_open

Non zero values of all the above metrics are exported over a one minute time interval. Each metric name is prefixed with hasura_cloud.

Graphs for all the above metrics can be viewed in your Datadog dashboard, under the Metrics tab. To navigate to the same, click View Metrics.

Datadog Integration successfully configured

Select the graphs you want to view from the metrics explorer. Alternatively, select the host you configured with this integration to see all the graphs corresponding to metrics exported by Hasura Cloud.

Metrics successfully exported to Datadog

View traces

Along with the logs and metrics, query tracing is exported to Datadog. From your Datadog metrics page, APM > Traces will show you a live search of ingested trace data, with each entry of the output table showing the duration of the operation.

Clicking on an entry from the table will show the details about the trace, including a flame graph, which contains a visual representation of where the operation spent its execution over time. For example, in this case a query took 2.31 milliseconds in total, of which 758 microseconds was the actual Postgres database processing the query.

Flame graph for a trace entry in Datadog