LivingDevOps outlines Kubernetes production playbook

LivingDevOps used a May 20 X thread to argue that basic Kubernetes tutorials do not reflect how teams run production systems, and tied that argument to an eight-week training course on Kubernetes, DevSecOps and AIOps. The post described a production stack that includes HashiCorp Vault and External Secrets Operator for secrets, Karpenter and KEDA for autoscaling, ArgoCD for multi-environment GitOps, and observability tools for incident detection. (livingdevops.com) LivingDevOps’ website lists an “8-Week real world Kubernetes on AWS Bootcamp with DevSecOps and AIOps” starting on May 23, with 32 live classes over eight weeks. The thread’s pitch was that platform and site-reliability engineers need operating patterns, not certification prep. (livingdevops.com) LivingDevOps’ course page uses similar language, saying it teaches “real production implementations” and “the way it actually works in companies.” ### Why did the post say tutorials are not enough? LivingDevOps said production Kubernetes work starts after pods are running, with teams managing secrets, scaling, deployments, pipelines and incidents across environments. The course outline on the company’s site backs that up with modules on probes, quotas, Pod Disruption Budgets, GitOps, CI/CD, monitoring and troubleshooting. (livingdevops.com) Akhilesh Mishra, named by LivingDevOps as the instructor on its courses page, frames the broader teaching approach in a separate April 13 post titled “Kubernetes Isn’t Hard. You’re Learning It Backwards.” In that article, he wrote that Kubernetes concepts make more sense when taught as responses to operational problems rather than as isolated definitions. (livingdevops.com) ### How does the secrets setup work in practice? HashiCorp documents Vault Secrets Operator as a supported way for pods to consume Vault secrets through Kubernetes-native resources. The LivingDevOps thread instead cited External Secrets Operator, or ESO, alongside Vault, a pattern widely used to keep secret values out of Git while syncing references from an external store into Kubernetes. (livingdevops.com) External Secrets Operator documentation from third-party implementation guides shows the same model: Git stores references and policy, while Vault or another backend remains the system of record for secret material. That matches the thread’s emphasis on production-safe secret handling rather than embedding credentials in manifests. (livingdevops.com) ### Why pair Karpenter with KEDA? LivingDevOps grouped Karpenter and KEDA together under autoscaling, reflecting two different layers of scaling. Karpenter adds or removes compute capacity for Kubernetes clusters, while KEDA is commonly used to scale workloads based on event sources such as queues or metrics pipelines; the pairing suggests a setup where infrastructure and applications scale independently. (developer.hashicorp.com) That is an inference from the tools named in the thread and how they are typically deployed. The company’s bootcamp outline separately lists Horizontal Pod Autoscaler, Vertical Pod Autoscaler and production EKS infrastructure, showing that scaling is taught as more than a single controller or metric. (oneuptime.com) ### What does the GitOps and DevSecOps part include? ArgoCD appears repeatedly in LivingDevOps materials as the GitOps layer for deployments. The eight-week course page lists GitOps from the first week and later adds multiple environments through Terraform workspaces for dev, staging and production, along with Helm packaging and CI/CD pipelines. (livingdevops.com) The same outline also includes image builds, deployments, migrations, AWS IAM roles for service accounts, and live troubleshooting. Those details align with the thread’s claim that production work combines application delivery, cloud permissions, release workflows and operational debugging. (livingdevops.com) ### Where does observability and AIOps fit? LivingDevOps said observability and AIOps are part of the production stack, and its course materials list Prometheus, Grafana and troubleshooting as core topics. A separate nine-week Kubernetes course page on the same site says it includes “Advanced AIops implementation,” suggesting the company is packaging monitoring and AI-assisted operations as part of a broader training offering. (livingdevops.com) The eight-week course page says classes run Tuesday through Friday for two hours each, for a total of 32 sessions. LivingDevOps’ courses page lists the start date as May 23, 2026, and names Akhilesh Mishra as instructor. (livingdevops.com)