observability.md

The observability stack offers monitoring solution designed to operate natively within Kubernetes clusters.

It consists of the following key components:

• Prometheus: A time-series database and monitoring system that stores metrics from various sources, including Kubernetes components, applications, and custom exporters.
• Alertmanager: A component responsible for handling alerts generated by Prometheus based on predefined rules, enabling notifications and other procedures.
• Grafana: A powerful data visualization and dashboarding tool that integrates seamlessly with Prometheus, allowing users to create customized dashboards and visualizations for monitoring AI services and resources.

For visual assistance, refer to the following Cluster observability dashboard

For visual assistance, refer to the following for Habana observability dashboard

For visual assistance, refer to the following for Models observability dashboard

For visual assistance, refer to the following for Cluster wide observability dashboard

To access the observability dashboard, follow these steps:

Initiate your web browser and proceed to navigate to the specified URL:

   https://<cluster-url>/observability/login
   #The cluster URL was configured during deployment in the cluster_url field  in inference-config.cfg file)
   # Please use the default code "prom-operator" and kindly change it upon your first login.

Further API Documentation is coverd under docs/api-spec.yml