Ultimate DevOps Project: AWS, Kubernetes, CI/CD with ArgoCD

Provision an AWS EKS cluster with EBS and EFS storage, deploy Ghost via Argo CD from GitHub, and observe through Prometheus, Grafana, Splunk, and Fluent Bit.

Install chocolatey, helm, kubectl, and aws cli to provision an EKS cluster; configure IAM user, and set up EBS and EFS storage with kubeconfig for local access.

Create EFS and EBS storage classes for the EKS cluster, configure the EFS file system and security groups, and establish Argo and app namespaces for GitOps deployment.

Explore how Argo CD enables GitOps for Kubernetes with git-driven continuous delivery, automated deployments, state syncing, and easy rollbacks.

Install and configure Argo CD on EKS using Bitnami helm chart, set storage class to gp2, expose via a load balancer, and use the UI as the GitOps control center.

Create a new public GitHub repository, generate a personal access token, and push your code via git with a Windows credential manager, preparing for GitOps with Argo CD.

Configure the Ghost helm chart to your needs by editing the values YAML, understanding templates, and mapping storage and MySQL dependencies, then expose Ghost with an AWS load balancer.

Deploy a Ghost blog with a Helm chart using Argo CD, enabling git-synced, self-healing deployments and monitoring Ghost MySQL metrics with Prometheus, Grafana, and Slack alerts.

Map your Ghost app running on AWS ELB to a custom domain using No-ip’s free subdomain, create a CNAME pointing to the Ghost ELB, and test DNS propagation.

Explore Ghost features, access the admin dashboard, create posts with the editor, collaborate with a team, customize themes, and optimize SEO with analytics and email newsletters.

Install and configure Prometheus on EKS to scrape MySQL metrics on port 9104 using a Bitnami Prometheus helm chart, with Grafana, Argo CD, and GitOps integration.

Prepare and deploy Grafana on Kubernetes using a helm chart, configure persistence with AWS EBS, expose via load balancer on port 80, and integrate with Argo CD for monitoring dashboards.

Learn how to connect Grafana to Prometheus inside a Kubernetes namespace, configure a Prometheus data source, test the connection, and import a MySQL overview dashboard to visualize live MySQL metrics.

Learn to integrate Grafana with Slack for real-time alerts by creating a Slack app, obtaining a bot token, and testing a Grafana contact point in the DevOps SRE project channel.

Set up a real-time alert in Grafana to detect drop table events in MySQL and notify via Slack, enabling proactive observability and defensive response.

Deploy fluentbit with a customized helm chart, mount the ghost pvc as a shared log directory, and configure it to stream ghost logs to splunk via the splunk hec endpoint.

Deploy Splunk in Kubernetes by using Helm charts for the Splunk operator and Splunk Enterprise, configure GP2 CSI storage with EBS, and expose a public load balancer for user access.

Deploy Splunk with Argo CD by creating two apps for Splunk operator and Splunk Enterprise. Fix crd payloads with server-side apply and fetch the auto-generated admin password from secrets.

Explore the Splunk search and reporting tab, and configure fluentbit to send logs to the main index via HTTP Event Collector tokens for monitoring and investigations.

Update fluentbit to stream ghost logs from the shared NFS volume to Splunk using a tail input, ghost tag, Splunk output with TLS and token, and a JSON filter.

Analyze json logs in Splunk to search Ghost CMS post updates, using automatic field extraction, filtering http put requests with status 200 to confirm health.

Explore advanced Splunk queries against ghost logs shipped by Fluentbit to classify errors with eval on status codes 400+ and create dashboards and alerts for application traffic.

Build a Splunk dashboard named ghost application health dashboard with a status-code pie chart (200, 400, 404, 500) and a top URLs table to monitor application health and spot anomalies.

Configure a dedicated Splunk alerts Slack app and incoming webhook to send real-time alerts to the DevOps SRE project channel, then restart the Splunk pod to apply changes.

Create actionable Splunk alerts to detect spikes in 400, 4 or 500 errors and login failures, and send real-time Slack notifications for rapid investigation.

Decommission the environment by cleaning up Argo CD apps, PVC data, Helm remnants, and resources; verify with kubectl, remove AWS volumes, and delete the EKS cluster to prevent cost leaks.