Edge Delta Developer Quickstart

From a developer’s standpoint, logs might trigger some eye-rolling emotions. They clog up your IDE with very verbose logs when developing something, and then for some odd reason, e.g business controls, you don’t have access to production logs when something goes wrong. Fast forward to modern times, the adage that slowness is the new down is true. With distributed systems and user journeys becoming commonplace, the science and discipline of observability are helping build more robust and performant services. For those who are not focused on observability, logs might be an eponym for any metric that is persisted, anything from spans to traces and even those good ol’ fashioned structured logs. Edge Delta aims to help simplify and scale the capturing and analysis of these crucial pieces of data and also help remediate issues. In this example, we will get started with Edge Delta’s solution and check out what insights we can garner from a few Kubernetes workloads.

Video:

https://www.youtube.com/watch?v=0bPVtoh-RqY&feature=youtu.be

Getting Started with Edge Delta

A few pieces to get started with Edge Delta, you can sign up for a free trial account on their SaaS.

Once you are signed up, log into the Edge Delta SaaS now you can start to connect your workload, logging, monitoring, or observability postures and solutions to Edge Delta.

To sandbox the example, let’s run some local workloads without any solution attached to the workload. This will give a good feel connecting Edge Delta to your environments.

A Sample Kubernetes Workload

Leveraging a local Kubernetes cluster such as minikube or k3d is a good way to run an example or your very own workload. In this example, let’s leverage minikube. If you don’t have minikube, on a Windows machine, you can use Chocolatey to install, or if using a Mac, Homebrew.

#Install Minikube

choco install minikube

brew install minikube

#Start Minikube and Validate

minikube config set memory 8128

minikube start

kubectl get pods -A

Once you get your local cluster up and running, you can deploy a sample application.

kubectl create deployment balanced — image=k8s.gcr.io/echoserver:1.4

kubectl expose deployment balanced — type=LoadBalancer — port=8080

To generate some logging details for the application, hit the application on the exposed external IP. In a new terminal window, can run the tunnel command for minikube.

minikube tunnel

Then get the external-ip.

kubectl get services balanced

Head to port 8080 of the external ip.

e.g http://10.105.141.99:8080/

Now you have generated some data, let’s wire this to Edge Delta.

Connecting Your Kubernetes Cluster to Edge Delta

To get started with Edge Delta, the first item is to create a Configuration to represent the Kubernetes Agent.

Edge Delta -> Data Pipeline -> Agent Settings + Create Configuration

In the Configuration Wizard, create a template with responsible defaults Kubernetes.

Click Create, and now you are ready to wire in the Agent. Clicking on the new Configuration that was just created, click on the green ellipses for Deploy Instructions.

Installing via Helm is the quickest route. If you don’t have Helm, you can install via your favorite package manager.

#Install Helm

choco install kubernetes-helm

brew install helm

Next wire in the Helm Repo. The instructions are also provided in the Deploy Configuration UI.

#Install Edge Delta

helm repo add edgedelta https://edgedelta.github.io/charts

helm repo update

After a few moments, you can head back to Edge Delta.

Diving in with Edge Delta

Back in the Edge Delta SaaS, after a few moments of collecting data, you can see what has been captured and starting to be analyzed. On the Observability Overview, you can start to see log data streaming in.

Edge Delta -> Observability-> Overview

Digging into more specific Kubernetes information can look at the K8s overview.

Edge Delta -> Observability -> K8s Overview

Digging in a little can visualize some of the traffic.

This is a short-lived application, and a majority of the analysis that Edge Delta can provide might not appear in the 5–10 minute demo. This is just the tip of the capabilities with Edge Delta.

Looking Forward with Edge Delta

Like any observability toolset, there are certainly a lot of options to take a look at. From very established vendors in the space to even cloud vendors trying to cannibalize everything in their ecosystem. Take Edge Delta for a spin yourself. Take a look at their documentation to get a better feel of what Edge Delta would look like in your environment.

Cheers,

-Ravi