An opinionated roadmap to become a Platform Engineer.
The roadmap is inspired by Teivah's SRE roadmap, but my idea is a bit different. In time, I want to create something similar to roadmap.sh but for Platform Engineers.
This picture is from my DevOpsDay Amsterdam 2024 talk - Platform Engineering's Inferno - you can find a video here.
I want this roadmap to be useful to the largest amount of people and if you want to help you are very welcome. Any contribution matters, may that be an input, a feedback or a PR.
Note: This repo also contains a Platform Engineering Manifesto which is playfully inspired by the Agile Manifesto.
- YAML (not even kidding you have to learn YAML. It's everywhere and it is the first thing to learn)
- Programming languages (Python, Golang, Typescript) - you don't necessarily need to be a master at these but it helps a ton
- Version control (Git)
- SDKs
- Bash scripting
- Navigating the filesystem
- Understand and manipulate both RAM and disk
- Understand and manipulate processes
- Understand cgroups and resource utilization
- Security (SELinux)
- Package management
- rip
- yum
- dnf
- Protocols
- TCP/IP vs UDP
- HTTP (http1 vs http2)
- TLS/SSL
- HTTPS
- DNS
- SSH
- Navigate and troubleshoot network
- DNS
- DHCP
- Firewall (Iptables)
- ISO-OSI model
- Network topologies
- Discoverability
- Load Balacing
- Proxy and reverse-proxy
- Queues
- Kafka
- RabbitMQ
- SQS
- ActiveMQ
- Event Log vs. Message Queue
- Load balancer
- Load distribution concepts
- Layer 4 vs. layer 7 load balancer
- APIs
- Stateful vs. stateless
- REST
- Microservices: pros and cons
- GraphQL
- gRPC
- Caching and key-value stores
- etcd
- Valkey / Redis
- Memcached
- In-memory databases
- CDN (Content Delivery Network)
- Cloudflare
- Akamai
- Fastly
- Databases
- Types of databases
- NoSQL vs. Relational databases
- MongoDB
- Relational vs. Document-based
- Column-oriented databases (OLAP)
- ClickHouse
- Graph databases
- Vector databases
- Objects-based storage
- NoSQL vs. Relational databases
- ACID principles
- CAP Theorem
- Indexing
- Transaction
- Storage
- Types of databases
- DevOps Concepts
- Culture over tools
- Feedback loops
- Effective communication
- Continuous improvement
- Toolchain
- Pipelines
- Automated testing
- Everything as -Ops (NetOps, DBOps, DocOps...)
- Cloud
- SaaS
- PaaS
- IaaS
- Serverless
- Knative
- Vendor-specific services to deploy serverless apps
- Environments
- Purpose of environments
- Isolation
- Multitenancy vs single tenancy
- CI/CD
- Vendor-specific services offering CI/CD capabilities
- SOPS
- Cloud Specific tools
- Hashicorp Vault
- GitOps
- ArgoCD
- FluxCD
- Temporal.io
- IaC (Infrastructure as Code)
- Configuration management
- Ansible
- Puppet
- Chef
- Provisioning and management
- Open-Tofu
- Terraform
- Pulumi
- CloudFormation (AWS proprietary)
- Bicep (Azure proprietary)
- Configuration management
- Cloud Providers
- AWS
- GCP
- Azure
- Version Control Products
- GitHub
- GitLab
- Bitbucket
- Concepts
- Bake time
- Feature flag
- Feature freeze
- Rollout supervision
- Rollout types
- Blue green rollout
- Canary rollout
- Progressive rollout
- Shadow rollout
- Concepts
- Difference between telemetry, monitoring and observability
- Trace vs. metric vs. log
- APM (Application Performance Monitoring)
- Percentiles
- Alerting
- ChatOps
- Slow vs. fast burn alert
- Log Management
- Levels
- Rotation
- Aggregation
- Retention
- Tools
- Prometheus
- Grafana
- OpenTelemetry
- Datadog
- New Relic
- Jaeger
- Sumo Logic
- Fluentd
- OpsGenie
- PagerDuty
- Encryption
- Algorithms
- Certificates
- TLS
- PKI
- Signatures
- Authentication
- Authorization
- IAM (Identity Access Management)
- OPA (Open Policy Agent)
- DevSecOps
- SAST
- DAST
- Container scanning
- Threat detection
- CNAPP (Cloud Native Application Protection Platform)
- CDR
- Concepts
- Hard vs. soft dependencies in a system
- Scalability
- Concepts
- Requests
- Rate limiting
- Throughput
- Practices
- Chaos engineering
- Containers
- Docker
- Tags
- OCI (Open Container Initiative)
- Container registry
- Container runtime
- Containerd
- CRI-O (Container Runtime Interface - Orchestrator)
- Container orchestration
- Kubernetes
- Nodes
- Control plane
- etcd
- kubectl
- Namespaces
- Pods
- Scheduler
- Health checks
- Liveness probes vs readiness probes
- Replica sets
- Deployments
- Services
- Ingress
- ConfigMaps
- Helm vs Kustomize
- Secrets
- DaemonSets
- Init containers & sidecar pattern
- eBPF and Kernel-level software
- StatefulSets
- Storage classes
- Persistent volumes and claims
- CSI (Container Storage Interface)
- CNI (Container Network Interface)
- Cilium
- ClusterIP vs. NodePort vs. LoadBalancer
- Service Meshes
- Linkerd
- Istio
- Network policies
- Service accounts
- RBAC (Role-based access control)
- Jobs & CronJobs
- HPA (Horizontal Pod Autoscaler)
- VPA (Vertical Pod Autoscaler)
- Cluster Autoscaler
- Node scaling
- Karpenter
- Pod Disruption Budgets
- Resource quotas
- Limit ranges
- Taints and tolerations
- Node affinity
- Pod affinity vs pod anti-affinity
- Pod security policies
- Admission controllers
- Operators
- CRDs (Custom Resource Definitions)
- CNCF Landscape
- Concepts
- Internal Development Platform
- Platform Capability
- Platform Catalog
- Golden Paths
- Self-service Platforms
- Platform as a Product
- DevEx (Developer Experience or UX for developers)
- OMF (Platform Market Fit)
- Platform ROI (Return On Investment)
- Internal Developer Platforms
- Backstage the OSS version is a builder of IDPs, not and IDP per se
- Mia Platform
- Humanitec
- Port
- Portainer
- Upbound
- Northflank
- Qovery
- Being a nice person to work with (no, really it's important)
- Communication
- Written
- Oral
- Presentation skills
- Collaboration
- Problem solving
- Curiosity
- Navigating ambiguity
- Blameless retrospectives
For a company to be successful in the long run, it needs to have a strong engineering culture. This is a list of things that can help to build such a culture.
