Author: Niha Tahoor M , Ayana Rukasar, Jitendra Singh, Rishika Kedia, Holger Wolf

The Red Hat® OpenShift® Cluster Observability Operator (COO) is generally available. COO is intended to enhance Red Hat OpenShift monitoring on IBM Z® and IBM® LinuxONE with HTTPS support, authentication, and metrics collection, along with defined alerting strategies.

The COO is a Red Hat OpenShift operator that enables the deployment of standalone, independently configurable monitoring stacks for different services and users. COO operator deploys three monitoring components. They are Prometheus, Thanos Querier, and Alertmanager.

The COO works separately from the default monitoring system managed by the Cluster Monitoring Operator (CMO). Both can run at the same time without interference, allowing you to use a COO monitoring stack alongside the default Red Hat OpenShift monitoring setup.

Benefits of using COO:

A COO-managed monitoring stack is useful when the default platform monitoring from the Cluster Monitoring Operator (CMO) cannot fully meet your requirements.

Compared to core platform and user workload monitoring, a COO-deployed monitoring stack offers several advantages

·       With a COO-deployed monitoring stack, users can add more metrics, which cannot be done in core platform monitoring without losing support.

·       The COO enables creating separate monitoring stacks for each user namespace or team, allowing independent configurations. This includes setting up alerts, alert routing, and alert receivers, without relying on a shared stack.

·       Multiple monitoring stacks can be created, managed by the COO operator in large clusters, helping when there are more metrics to handle.

·       This offers various observability plugins for extended support and debugging, like as viewing dashboards, troubleshooting panel, distributed tracing UI plugin, and logging UI plugin. 

Observability UI Plugins: 

·       Dashboard UI plugin: The dashboard UI plugin in OpenShift lets you enhance dashboards in the web console under Observe → Dashboards. You can add extra Prometheus data sources from the cluster alongside the default one, creating a unified view of data from multiple sources.

Use case: We can manually create a dashboard of our choice from GitHub or any other open source and verify whether our dashboard is reflected or not.

·       Distributed tracing UI plugin: The distributed tracing UI plugin in OpenShift adds tracing features to the Administrator view under Observe → Traces. It helps track requests from the front end to the backend of microservices, making it easier to find code errors and performance issues in distributed systems.

Use case: When you have a Tempostack or TempoMonolithic instance in the cluster, we can observe the traces by setting a time range and selecting the stack. The traces can be displayed as a scatter plot, and the information, such as trace name, number of spans, and duration, can be seen underneath the traces.

·   

·       Logging UI plugin : The logging UI plugin helps in viewing the logging data in the web console on the Observe → Logs page.

Use case: In the logging scenario, after deploying the cluster logging with the help of this feature, we can visualise various types of logs.

·       Troubleshooting UI plugin :  The troubleshooting UI plugin in OpenShift gives a signal correlation, which is available under Observe → Alerting. With the help of this panel, you can correlate metrics, logs, alerts, and netflows with various data sources

Use case: the troubleshooting panel has query details and the topology graphs of the

Installation steps:

·       In the web console, go to the operator hub and search for observability in the search tab, click on the Install option.

·       After the operator is installed, you can check if the pods are up and running

Example use case with Dashboard UI plugin:

·       Install the operator following the above steps. Now create the UI plugin of the type dashboards by applying the below YAML

·       Check for the UI pod in the namespace

·       Create a datasource, apply the below YAML file.

·       Now we need to create a dashboard, we can take any open-source Prometheus supporting JSON file dashboard template and apply to the cluster like the below example

·       Download the below config file (note : this is just a reference) and run the steps below to support the dashboard extension

·       Once done we can see the plugin_test, overview, and related dashboards

Summary

With the availability of the Cluster Observability Operator on IBM Z and IBM® LinuxONE, administrators are enabled to create standalone monitoring stacks that are independently configurable for use by different services and users, complementing the built-in monitoring capabilities of Red Hat OpenShift.

Reference:

           https://docs.redhat.com/en/documentation/openshift_container_platform/4.18/html/cluster_observability_operator/index