本文へスキップ

このページの主な内容は英語です。このページを翻訳

キャリアロードマップ

Cloud Engineer roadmap

Cloud engineers build and operate the infrastructure that apps run on: networks, servers, containers, databases, security and automation. Certifications help here more than in most tech careers.

時間
⏱ 5–7 months
Level
Intermediate
Steps
6 steps
Architecture ↓Cheat sheet ↓5 interview Q&As ↓
共有

The roadmap

Click a step to open the full guide.

Best architecture: a highly available cloud deployment

The classic multi-zone setup for production workloads.

Multi-AZ cloud architecture

  1. DNS + CDN + WAF
  2. Load balancer (public subnets, 2+ zones)
  3. App containers / VMs (private subnets, autoscaled)
  4. Managed DB (multi-AZ) · cache · object storage
  5. IaC · CI/CD · monitoring · backups

Why it's built this way

  • Spread across two or more availability zones.
  • Only the load balancer is public; everything else is private.
  • Managed, multi-AZ databases with automated backups.
  • Everything defined in infrastructure as code and deployed by CI/CD.

Cloud Engineer cheat sheet

The essentials on one screen. Bookmark it for revision.

Linux & network

Disk usagedf -h / du -sh *
Processestop / ps aux | grep app
Service logsjournalctl -u app -f
DNS lookupdig example.com
HTTP debugcurl -v https://…
Open portsss -tulpn

Terraform

Initterraform init
Previewterraform plan
Applyterraform apply
Formatterraform fmt
Destroy (careful)terraform destroy

Improve performance

  • Use autoscaling and load balancers across zones.
  • Put a CDN in front of static content and APIs.
  • Choose regions close to users; right-size instances.
  • Cache in memory (Redis) and use read replicas.
  • Use managed services to reduce operational overhead.

Security checklist

  • MFA and least-privilege IAM; no shared or root credentials.
  • Private subnets for servers and databases; minimal open ports.
  • Encrypt data at rest and in transit.
  • Centralised logging, audit trails and threat detection.
  • Secrets in a secrets manager; patch images and hosts.
  • Infrastructure as code with peer review for every change.

Most important interview questions

Click a question to see a strong answer. Each step also has its own questions.

1the shared responsibility modelとは？

The cloud provider secures the infrastructure; you secure your data, identities, configuration and applications.

2Explain a VPC with public and private subnets.

Public subnets route to an internet gateway (for load balancers); private subnets have no inbound internet access (for apps and databases) and reach out via NAT.

3What are RTO and RPO?

RTO is the maximum acceptable downtime; RPO is the maximum acceptable data loss measured in time.

4Blue-green vs canary deployment?

Blue-green switches all traffic between two full environments; canary shifts a small percentage first and increases gradually.

5How do you secure secrets in the cloud?

Store them in a managed secrets service, grant access via roles, rotate regularly, never commit them, and audit access.

Related件のロードマップ

6

AWS

6 steps · 6–8 weeks · Intermediate

The most widely used cloud platform.

View roadmap →
5

Docker & Kubernetes

5 steps · 4–6 weeks · Intermediate

Package apps and run them anywhere.

View roadmap →
6

Backend Developer

6 steps · 5–6 months · Intermediate

Build the APIs, databases and services behind apps.

View roadmap →
1

Cloud Engineer · Step 1 of 6 · 3 weeks

Linux and the command line

The operating system of the cloud.

Almost every server in the cloud runs Linux. Be comfortable moving around, editing files, managing services and reading logs from a terminal.

Scripting in Bash or Python automates repetitive admin tasks.

What you'll learn

  • Files, permissions and users
  • Processes and services
  • Package managers
  • SSH
  • Bash scripting
  • Logs

Key Linux folders

  • /
  • /etc (config)
  • /var/log (logs)
  • /home (users)
  • /usr/bin (programs)
  • /tmp (temporary)

Key concepts

Permissions (rwx)
Read, write, execute for owner, group, others.
Process
A running program with its own ID (PID).
SSH
Secure remote login using keys.

Step by step

  1. 1.1

    Navigate

    ls, cd, pwd, find, grep, pipes |.

  2. 1.2

    Permissions

    Users, groups, chmod, chown, sudo.

  3. 1.3

    Services and logs

    systemctl, journalctl, tail -f.

  4. 1.4

    Script

    Automate a daily task with Bash.

Notes

  • Learn grep, find, tail -f and systemctl early.

Tips & tricks

  • Use Linux as your daily system for a month, or run it in a VM.

Interview questions

How do you find what's using all the disk space?

df -h to see full filesystems, then du -sh /* (and deeper) to find large folders; check logs in /var/log and old Docker images.

The quick way to study this

Set up a Linux VM and practise one admin task every day.

2

Cloud Engineer · Step 2 of 6 · 2 weeks

Networking fundamentals

How traffic flows between systems.

Cloud networking (VPCs, subnets, security groups) is built on these basics. Most outages and 'it cannot connect' issues are networking problems.

Learn to subnet confidently; it appears in every cloud certification.

What you'll learn

  • IP addresses and subnets (CIDR)
  • DNS
  • HTTP/HTTPS and TLS
  • Ports and firewalls
  • Load balancers
  • VPNs

DNS resolution

  1. Browser asks
  2. Resolver
  3. Root server
  4. TLD (.com)
  5. Authoritative server
  6. IP returned

Key concepts

CIDR
/16 = first 16 bits fixed → 65,536 addresses; /24 → 256.
DNS
The internet's phone book: names → IP addresses.
Port
A numbered door on a machine: 22 SSH, 80 HTTP, 443 HTTPS, 5432 PostgreSQL.

Step by step

  1. 2.1

    IP and CIDR

    Understand 10.0.0.0/16 and how to split it into subnets.

  2. 2.2

    DNS

    A, CNAME, MX records; how names resolve.

  3. 2.3

    TLS

    Certificates and HTTPS.

  4. 2.4

    Debug

    dig, curl -v, ping, traceroute.

Notes

  • Private subnets for databases, public subnets only for load balancers.

Tips & tricks

  • Use dig, curl -v and traceroute to debug connections.

Interview questions

How many addresses are in a /24?

256 (2^8). In cloud subnets a few are reserved by the provider, so slightly fewer are usable.

The quick way to study this

Learn CIDR with practice problems, then draw the network of a simple web app.

3

Cloud Engineer · Step 3 of 6 · 5 weeks

Core cloud services

Compute, storage, databases and IAM on one cloud.

Pick one cloud and learn its core services well. AWS has the largest market share; Azure is strong in enterprises; Google Cloud in data and AI.

Set up billing alerts on day one so practice never surprises you with a bill.

What you'll learn

  • Identity and access management
  • Virtual machines
  • Object storage
  • Managed databases
  • VPC networking
  • Billing and cost alerts

Cloud service models

IaaS

  • Virtual machines
  • You manage OS & app

PaaS

  • Managed platforms
  • You manage the app

SaaS

  • Ready software
  • You just use it

Key concepts

IaaS / PaaS / SaaS
You manage: servers / just your app / nothing (you use software).
Availability zone
A separate data centre; use 2+ for resilience.

Step by step

  1. 3.1

    Identity first

    IAM, MFA, least privilege, budgets.

  2. 3.2

    Network

    VPC with public/private subnets across two zones.

  3. 3.3

    Compute, storage, database

    VMs, object storage, managed database.

  4. 3.4

    Certify

    Take a foundational or associate certification.

Notes

  • Least privilege: give each user and service only the permissions it needs.

Tips & tricks

  • Earn a foundational certification to structure your learning.

Interview questions

How do you make an application highly available in the cloud?

Run it across multiple availability zones behind a load balancer, use a managed multi-AZ database, autoscaling, health checks and backups.

The quick way to study this

Follow the AWS roadmap and take the Cloud Practitioner or Solutions Architect Associate exam.

AWS roadmap →
4

Cloud Engineer · Step 4 of 6 · 3 weeks

Infrastructure as code

Build infrastructure from version-controlled files.

Clicking in consoles does not scale. Infrastructure as code describes your setup in files you can review, version and repeat.

Terraform-style tools work across clouds and are widely requested in job posts.

What you'll learn

  • Terraform or OpenTofu
  • Cloud-native IaC (CloudFormation, Bicep)
  • State management
  • Modules
  • Reviewing changes with plans

Terraform workflow

  1. Write .tf
  2. terraform init
  3. terraform plan
  4. Review
  5. terraform apply

Key concepts

State file
Terraform's record of what it created; keep it safe and remote.
Drift
Real infrastructure differs from the code (e.g. manual console changes).
Idempotent
Applying the same config again makes no changes.

Step by step

  1. 4.1

    Write config

    Define resources in Terraform/OpenTofu files.

  2. 4.2

    Plan and apply

    Review the plan, then apply.

  3. 4.3

    Remote state

    Store state remotely with locking.

  4. 4.4

    Modules

    Reuse patterns across environments (dev, prod).

Notes

  • Always review the plan before applying.

Tips & tricks

  • Rebuild your console-made projects in code.

Interview questions

Why use infrastructure as code?

Repeatable, reviewable, version-controlled infrastructure; fewer manual mistakes; easy to recreate environments and audit changes.

The quick way to study this

Write Terraform for a VPC, a VM and a database, then destroy and recreate it.

5

Cloud Engineer · Step 5 of 6 · 3 weeks

Containers and Kubernetes

Package and orchestrate applications.

Containers are the standard way to ship apps. Kubernetes runs them at scale but adds complexity; managed services and serverless containers are often simpler.

Learn when Kubernetes is worth it and when a simpler service is better.

What you'll learn

  • Docker images and containers
  • Container registries
  • Kubernetes basics
  • Managed Kubernetes
  • Helm
  • Serverless containers

Where to run containers

Serverless containers

  • Least ops
  • Scale to zero
  • Simple apps

Managed Kubernetes

  • Most control
  • Many services
  • Needs expertise

Key concepts

Container registry
Storage for container images.
Orchestration
Scheduling, healing and scaling containers across machines.

Step by step

  1. 5.1

    Containers

    Build images; push to a registry.

  2. 5.2

    Managed runtimes

    Serverless containers first; Kubernetes when needed.

  3. 5.3

    Kubernetes basics

    Deployments, services, ingress, autoscaling.

Notes

  • Small images build and deploy faster.

Tips & tricks

  • Run a local Kubernetes cluster to practise.

Interview questions

When would you choose Kubernetes?

Many services, need for portability, fine-grained control over networking/scaling, and a team able to operate it; otherwise managed container services are simpler.

The quick way to study this

Containerise an app and deploy it to a managed container service.

Docker & Kubernetes roadmap →
6

Cloud Engineer · Step 6 of 6 · 3 weeks

CI/CD, monitoring and security

Automate delivery and keep systems healthy.

Cloud engineers keep systems reliable, secure and affordable. Automate deployments, monitor everything and test your backups.

Cost optimisation is a valued skill: right-sizing, auto-scaling and shutting down unused resources.

What you'll learn

  • CI/CD pipelines
  • Monitoring and alerting
  • Logging
  • Backups and disaster recovery
  • Security best practices
  • Cost optimisation

Operate in the cloud

  1. Deploy via CI/CD
  2. Monitor & alert
  3. Respond to incidents
  4. Review costs
  5. Improve

Repeats as a loop

Key concepts

RTO / RPO
RTO: how fast you must recover. RPO: how much data you can lose.
FinOps
Managing cloud costs as a shared team practice.

Step by step

  1. 6.1

    CI/CD

    Pipelines for infrastructure and applications.

  2. 6.2

    Observe

    Dashboards, alerts, log search.

  3. 6.3

    Backups & DR

    Automated backups, tested restores, recovery plan.

  4. 6.4

    Cost

    Tags, budgets, right-sizing, scheduling.

Notes

  • A backup you have never restored is not a backup.

Tips & tricks

  • Build a portfolio project with Terraform, CI/CD and dashboards, documented in a README.

Interview questions

How would you cut a cloud bill by 30%?

Find top costs, delete idle resources, right-size instances, autoscale and schedule non-prod shutdowns, buy commitments for steady loads, use spot for flexible jobs, and add storage lifecycle rules.

The quick way to study this

Add a GitHub Actions pipeline that deploys your Terraform and app automatically.

このページを共有