Integrated monitoring in SMM

One of the core features of the Istio service mesh is the observability of network traffic. Because all service-to-service communication is going through Envoy proxies, and Istio’s control plane is able to gather logs and metrics from these proxies, the service mesh can give you deep insights about your network.

But Istio needs some additional components to unleash its full potential of observing the mesh. Prometheus collects these metrics from Envoy proxies, and Grafana displays monitoring information on analytics dashboards.

Service Mesh Manager builds an integrated, production-ready environment of these components with a single CLI command. Prometheus is set up to scrape Envoys, and Grafana dashboards are automatically configured for services and workloads in the mesh.

Note: Service Mesh Manager provides an end-to-end monitoring solution built on Service Level Objectives. To start using this feature, see Tracking Service Level Objectives (SLOs).

Test integrated monitoring

The MENU > TOPOLOGY and MENU > SERVICES pages of the Service Mesh Manager UI serve as starting points to diagnose problems within the mesh. You’ll see if error rates are up for a specific service, RPS is down, or latency increases.

Some Grafana dashboards are available directly in the Service view, but for more in-depths analytics, you’ll probably need to dig deeper into Grafana dashboards, or directly into Prometheus metrics. To access a specific Grafana dashboard from Service Mesh Manager, just click the Grafana link (Open metrics in Grafana ) in the Service or Workload view.

To automate monitoring, you can configure Service Level Objectives (SLOs) and alerts as well. For details, see Tracking Service Level Objectives (SLOs).

To access the Service Mesh Manager UI from your machine, use the smm dashboard command.

Grafana integration Grafana integration

Advanced use-cases

  1. Grafana and Prometheus are installed in the smm-system namespace and are available on an internal ingress gateway. When using smm dashboard, that internal ingress gateway is securely proxied from the cluster, so Service Mesh Manager will be accessible from localhost. Grafana and Prometheus is also available on the ingress gateway, and they are proxied on separate URIs. Grafana is accessible directly on http://127.0.0.1:50500/grafana, while Prometheus is available on http://127.0.0.1:50500/prometheus.

  2. Federating Istio metrics from the Prometheus instance of Service Mesh Manager. Currently it’s not possible to specify an external Prometheus instance to use when installing Service Mesh Manager. If you have other metrics on another Prometheus instance, and want to have them in one place, we suggest to set up federation between the two Prometheus instances.