

Observability Deployment Strategy¶

WSO2 Enterprise Integrator 7.1 offers two observability solutions for monitoring and managing your integration deployments in the Micro Integrator. You can choose one of the two solutions depending on the nature of your Micro Integrator deployment.

Choosing an observability strategy¶

The cloud-native solution is more suitable in the following scenarios:

You are creating a new cloud-native Micro Integrator deployment.

Note

See the instructions on setting up a cloud-native Micro Integrator deployment on Kubernetes.
You already have Prometheus, Grafana, and Jaeger as your in-house monitoring and observability tools. This applies to VM deployments as well as Kuberentes deployments.

The classic observability solution is more suitable in the following scenarios:

You require more business analytics and less operational observability.
You require a simpler deployment.
You already have an observability stack such as ELK.

Cloud-Native Observability¶

The following diagram depicts the complete cloud-native observability solution for your Micro Integrator deployment, which includes metrics monitoring, log monitoring, and message tracing capabilities.

Cloud Native Deployment Architecture

You can also set up different flavours of this solution depending on your requirement.

Technologies¶

The cloud-native observability solution is based on proven projects from the Cloud Native Computing Foundation, which makes the solution cloud native and future proof. Following are the technologies used in the current solution:

Feature	Technology
Metrics	Prometheus
Visualization	Grafana
Logging	Log4j2, Fluent-Bit, and Grafana Loki
Tracing	Jaeger

Minimum cloud-native observability¶

The basic deployment offers you metrics capabilities. You can set up the basic deployment with only Prometheus and Grafana to view and explore with the available Prometheus metrics.

Cloud Native Deployment - Minimum

Log processing add on¶

Once you set up the basic deployment, you can integrate log-processing capabilities. To use this, you need to install Fluent-Bit as the logging agent and Grafana Loki as the log aggregator.

Cloud Native Deployment with Logs

Message tracing add on¶

Once you set up the basic deployment, you can integrate message tracing capabilities. To use this you need to install Jaeger.

Cloud Native Deployment with Tracing

Classic Observability¶

In the solution, metrics monitoring and message tracing is handled using the EI Analytics. For log monitoring, use log4j2 logs. This is the same solution that was available for previous EI versions.

Technologies¶

Feature	Technology
Metrics and Tracing	EI Analytics server
Visualization	EI Analytics Portal
Logs	Log4j2

Metrics and message tracing¶

The following diagram depicts how EI Analytics is used for metrics monitoring and message tracing.

EI-Analytics Observability

Log processing¶

With the classical observability strategy, you cannot view logs from a dashboard (as you do with the cloud-native strategy). You need to first configure log4j2 logs for log monitoring. You can then use the Micro Integrator dashboard to download log files and manage them.

What's Next?¶

If you choose cloud-native observability, see the relevant topic for instructions:

Set up cloud-native observability on a VM.
Set up cloud-native observability on Kubernetes.

If you choose classic observability, set up EI Analytics for statistics and message tracing.

Top