DevOps & Infrastructure¶

Comprehensive reference covering Docker, Kubernetes, CI/CD, Terraform, cloud platforms, monitoring, SRE practices, and data center networking.

Containers & Docker¶

• docker fundamentals - containers, images, volumes, networking, lifecycle
• dockerfile and image building - Dockerfile syntax, multi-stage builds, optimization
• docker compose - multi-container orchestration, service discovery
• docker for ml - ML-specific Docker patterns, JupyterLab, MLflow, Model Runner

Kubernetes¶

• kubernetes architecture - control plane, worker nodes, CNI, cluster deployment
• kubernetes workloads - Pods, Deployments, StatefulSets, DaemonSets, ConfigMaps, Secrets
• kubernetes services and networking - Services, Ingress, TLS, cert-manager, DNS
• kubernetes storage - PV, PVC, StorageClass, CSI drivers, cloud storage
• kubernetes resource management - requests, limits, QoS, namespaces, quotas, HPA, autoscaling
• kubernetes on aks - Azure AKS, ACR, Azure AD, monitoring, virtual nodes
• kubernetes on eks - AWS EKS, ECR, EBS/EFS, ALB Ingress Controller

Package Management & Templating¶

• helm package manager - charts, templates, hooks, releases, repositories, secrets

CI/CD & Automation¶

• cicd pipelines - GitHub Actions, Azure DevOps, pipeline stages, multi-environment
• jenkins automation - Jenkinsfile, shared libraries, Docker agents, credentials
• gitops and argocd - ArgoCD, sync waves, app-of-apps, Sealed Secrets, Crossplane

Infrastructure as Code¶

• terraform iac - HCL, state, modules, workspaces, cloud providers
• ansible configuration management - playbooks, roles, inventory, idempotency

Cloud Platforms¶

• aws cloud fundamentals - IAM, VPC, EC2, S3, CloudWatch, ECR, App Runner
• container registries - Docker Hub, ECR, ACR, GAR, Nexus

Monitoring & Observability¶

• monitoring and observability - Golden Signals, Prometheus, Grafana, Loki, Tempo, SLI/SLO/SLA

SRE Practices¶

• sre principles - culture, error budgets, toil, team models
• sre incident management - on-call, postmortems, escalation, diagnostics
• sre automation and toil - automation maturity, tools, workflow automation
• chaos engineering and testing - chaos engineering, load testing, game days, resilience patterns

Deployment & Release¶

• deployment strategies - rolling update, blue-green, canary, feature flags, 15-factor methodology
• service mesh istio - Istio, Envoy, traffic management, mTLS, observability

Architecture Patterns¶

• microservices patterns - Spring Cloud, API Gateway, service discovery, event-driven, OAuth2
• datacenter network design - CLOS/leaf-spine, VXLAN, EVPN, BGP, multi-site

Foundations¶

• devops culture and sdlc - DevOps principles, Agile/Scrum, SDLC, tools landscape
• git version control - Git workflow, branching strategies, monorepo vs multirepo
• linux server administration - filesystem, processes, networking, SSH, systemd, scripting

Additional References¶

• comfyui container build - Patterns for building production-grade ComfyUI Docker images: distutils conflict resolution, layer
• container security scanning - Automated detection of vulnerabilities, misconfigurations, and secrets in container images, IaC
• kubernetes operators - Software extensions that use Custom Resources (CRs) to manage applications and their components
• kubernetes security - Multi-layered defense for Kubernetes clusters: authentication, authorization, admission control
• kustomize - Template-free customization of Kubernetes YAML configurations
• libvirt kvm networking - Diagnosing and fixing intermittent connection drops on KVM guests using virbr0 NAT
• observability query languages - Reference for the three main observability query languages - Prometheus PromQL for metrics, Grafana
• ois multi environment task monitoring architecture - The OIS infrastructure utilizes a uniform monorepo deployment across staging and production
• runpod flash gpu serverless - RunPod Flash is an open-source Python SDK (MIT license, v1.16.0) that enables the deployment of
• runpod production - Reference for deploying GPU workloads on RunPod: deploy type selection, image strategy tiers
• runpod serverless diagnostic image - Serverless workers die silently
• runpod serverless observability limits - RunPod Serverless workers are difficult to debug not because of one missing feature but because
• runpod serverless python silent exit 1 - Analysis and mitigation of silent worker crashes (exit code 1) during task pickup on RunPod
• runpod serverless silent worker crashes on task pickup - Workers in a serverless environment may spawn as "Ready," pick up a task, and then die silently
• runpod serverless stuck queue idle workers - Symptom: health endpoint shows idle > 0 or ready > 0, queue depth is positive, inProgress = 0, and
• tunnel architecture - Exposing VM-hosted or NAT-ed services to the public internet without opening inbound firewall ports