Coroot - The Next Level in Kubernetes ObservabilityTo follow up on my previous Kubernetes articles:

I would like to introduce a project which brings Kubernetes observability to the next level and provides major help in visualizing and diagnosing Kubernetes deployments.

The project is Coroot and it comes with the Community version Coroot Community Edition (source code at which you can install on your Kubernetes. The installation using helm charts is trivial:

Enabling port forwarding to access Coroot through the web:

The first you can see is the graph representing Kubernetes deployment. Why is it helpful? Let’s review what we can see in the plain Kubernetes:

List of pods:

And the list of services:

I do not know how this is for you, but I have a hard time processing these lists, while Coroot offers visual help. Coroot is a web application, so there will be a lot of screenshots.

There is a glance overview of my deployments, including physical nodes:

MySQL Kubernetes deployments

And from there we can drill down into individual components, for example, cluster1-mysql:


And there you can highlight important relations, like for HAProxy:

HAProxy coroot

We can see that it handles pod mysql-1 and mysql-2 right now, even for the MySQL setup itself:

handles pod mysql-1 and mysql-2

We can see mysql-2 is the primary and mysql-0 and mysql-1 are replicas. There is also a network layout:

coroot kubernetes network layout

This shows network connections between different physical nodes and pods mapping to nodes. If we run a scenario with MySQL primary failure (similar to what we did in Using Percona Kubernetes Operators With K3s Part 2: Percona Server for MySQL Operator), it also can be visualized.

Node failure:

coroot node failure

Primary failover from mysql-0 to mysql-2:

primary fallover coroot

In general, Coroot offers visibility into a variety of problems you may encounter in Kubernetes, and there are more details on their Failurepedia page.


In my view, Coroot brings the visibility of Kubernetes deployments to the next level and is a must-have tool for the diagnostics and troubleshooting of Kubernetes workloads. If you run your application in Kubernetes, I recommend you give it a try!

Notify of

Inline Feedbacks
View all comments