OpenTelemetry for Mendix Cloud

Introduction

Mendix Cloud supports sending telemetry—traces, metrics, and logs—directly to any OpenTelemetry Protocol (OTLP)-compatible observability backend, such as Grafana Cloud.

When enabled, the Mendix operator automatically injects an OpenTelemetry Collector sidecar container alongside your application. This sidecar receives telemetry from the Mendix Runtime and forwards it to your chosen backend. No manual installation of the collector is required.

For more information about what metrics and data Mendix provides, see Monitoring Your Mendix Apps with an APM Tool.

Prerequisites

This integration exclusively supports destinations that ingest telemetry via standard core exporters. Compatible backends include any vendor or platform that natively accepts inbound OTLP (via gRPC or HTTP) or Prometheus Remote Write (PRW) streams. Custom, proprietary vendor SDK exporters are not supported.

Before connecting your app to an OpenTelemetry backend, ensure you have:

• Access to an OTLP-compatible observability backend (for example, Grafana Cloud)
• The OTLP ingestion endpoint URL for your backend
• An API token or other credentials for authenticating with the backend
• A licensed Mendix Cloud app

Connecting Your Node to an OpenTelemetry Backend

To connect your Mendix Cloud environment to an OpenTelemetry backend:

1. Go to Apps.

2. Click Environments on the app of choice.

3. Click Details ( ) for the environment you want to configure.

4. Go to the Runtime tab.

5. In the Custom Environment Variables section, add the following environment variables:

   Variable	Description	Default
   MX_OTEL_ENABLED	Master switch. Set to true to enable the integration.	false
   MX_OTEL_EXPORTER_OTLP_ENDPOINT	OTLP ingestion URL for all signals (traces, metrics, and logs).	—
   MX_OTEL_EXPORTER_OTLP_HEADERS	HTTP headers sent with every request. Used for authentication. Format: key1=value1,key2=value2.	—
   MX_OTEL_EXPORTER_OTLP_PROTOCOL	Transport protocol. One of grpc, http/protobuf, or http/json.	http/protobuf

6. Optionally, set service identity variables to control how your app appears in your observability backend. For more information, refer to the Service Identity section below.

7. Click Save.

Example – Grafana Cloud

Grafana Cloud exposes a single OTLP endpoint for all signals over HTTPS. To connect your app:

1. In your Grafana Cloud stack, navigate to Getting Started guide > OpenTelemetry > Serverless/Other.

2. Create an access policy token with metrics:write, logs:write, and traces:write scopes.

3. Copy the OTLP endpoint URL and your token.

4. Set the following environment variables in your Mendix environment:

   MX_OTEL_ENABLED=true
   MX_OTEL_EXPORTER_OTLP_ENDPOINT=https://otlp-gateway-prod-eu-west-0.grafana.net/otlp
   MX_OTEL_EXPORTER_OTLP_HEADERS=Authorization=Basic BASE64_ENCODED_INSTANCE_ID:TOKEN
   MX_OTEL_SERVICE_NAME=my-mendix-app  # only necessary if you want to override the default service name (app subdomain)
   MX_OTEL_DEPLOYMENT_ENV=production   # only necessary if you want to set the deployment environment (defaults to `none`) or can be set via the `env` application tag

   Replace BASE64_ENCODED_INSTANCE_ID:TOKEN with the base64 encoding of <instance-id>:<token> as shown on the Grafana Cloud OTLP configuration page.

5. Redeploy your app.

Within a minute of the first request to your app, traces appear in Grafana Cloud's Explore view, and metrics are visible in Prometheus datasource queries.

Additional Information

Per-Signal Configuration

By default, all three signals—traces, metrics, and logs—are sent to the backend configured with MX_OTEL_EXPORTER_OTLP_ENDPOINT. You can control each signal independently using the following variables:

When not set explicitly, the exporter for each signal is inferred from the endpoint variables that are present. Setting a signal to none disables it. The sidecar accepts the data from the Runtime but discards it without forwarding.

Per-Signal OTLP Overrides

If your backend requires different endpoints, headers, or protocols per signal, you can override the base OTLP configuration for each signal individually:

Log Redaction

Email addresses are automatically redacted from log entries before they are sent to your backend. To disable redaction, set LOGS_REDACTION to false. By default, this variable is set to true.

Also, certain secret patterns, such as database connection strings and storage endpoint URLs logged by the Mendix Runtime are always redacted regardless of this setting and cannot be turned off.

Prometheus Remote Write (Metrics Only)

If your metrics backend uses Prometheus Remote Write (for example, Grafana Mimir, Thanos, or VictoriaMetrics), you can send metrics via PRW while still sending traces and logs via OTLP.

OTel resource attributes (service name, version, environment, and custom tags) are automatically converted to Prometheus labels when using this exporter.

Example - metrics to a PRW endpoint, traces and logs via OTLP:

MX_OTEL_ENABLED=true

# Traces and logs via OTLP
MX_OTEL_EXPORTER_OTLP_ENDPOINT=https://otlp.example.com
MX_OTEL_EXPORTER_OTLP_HEADERS=Authorization=Bearer YOUR_OTLP_TOKEN

# Metrics via Prometheus Remote Write (takes precedence over OTLP for metrics)
MX_OTEL_METRICS_EXPORTER=prometheusremotewrite
MX_OTEL_EXPORTER_PROMETHEUSREMOTEWRITE_ENDPOINT=https://mimir.example.com/api/v1/push
MX_OTEL_EXPORTER_PROMETHEUSREMOTEWRITE_HEADERS=Authorization=Bearer YOUR_MIMIR_TOKEN

MX_OTEL_SERVICE_NAME=my-mendix-app

Service Identity

These variables control the resource attributes attached to all telemetry. They determine how your app is identified in your observability backend.

OpenTelemetry Issues

If you encounter issues with the OpenTelemetry integration that are not caused by your backend configuration, submit a support request through Mendix Support Portal.

Read More

• Metrics - Mendix Runtime