bongoDev

AWS is not just a collection of services. For a cloud engineer it is a programmable operating environment made of accounts, identities, APIs, networks, compute, data services, observability systems and automation boundaries.

The cloud-engineering loop

Text

REQUIREMENT
   ↓
SERVICE SELECTION
   ↓
IDENTITY + NETWORK + DATA BOUNDARIES
   ↓
AUTOMATION / IaC
   ↓
DEPLOY
   ↓
OBSERVE
   ↓
SECURE + SCALE + OPTIMIZE
   ↓
RECOVER / IMPROVE

A cloud engineer must be able to answer four questions for every workload:

Where does it run? Region, Availability Zones, VPC, subnets, compute platform.
Who can access it? IAM, roles, resource policies, network controls.
Where does data live? S3, EBS, EFS, RDS, DynamoDB, caches, backups.
How do we know it is healthy? Metrics, logs, traces, events, audit history and alarms.

Shared responsibility

AWS operates the cloud infrastructure. You remain responsible for how you configure and use cloud services. The exact boundary changes by service type.

Model	AWS manages more of	You manage more of
EC2	facilities, physical hardware, virtualization	guest OS, patches, app runtime, security groups, data
RDS	database infrastructure and many operational tasks	schema, users, queries, data, configuration choices
Lambda	servers and runtime infrastructure	function code, permissions, configuration, dependencies
S3	storage infrastructure	bucket policy, object permissions, encryption choices, lifecycle

Popular service selection map

Need	Common AWS choice
virtual machine	EC2
object storage	S3
block storage	EBS
shared file system	EFS
relational database	RDS / Aurora
NoSQL key-value/document	DynamoDB
cache	ElastiCache
HTTP load balancing	ALB
TCP/UDP load balancing	NLB
DNS	Route 53
CDN	CloudFront
containers	ECS / Fargate / EKS
functions	Lambda
messaging	SQS / SNS / EventBridge
metrics/logs/alarms	CloudWatch
API audit history	CloudTrail
remote server operations	Systems Manager

Six architecture lenses

AWS Well-Architected thinking can be reduced to questions cloud engineers ask daily:

Operational excellence: can we deploy, observe and recover predictably?
Security: is access least-privilege and data protected?
Reliability: what fails, how is it detected, and how do we recover?
Performance efficiency: are we using the right service/resource shape?
Cost optimization: are we paying for idle or oversized capacity?
Sustainability: are we avoiding unnecessary resources and waste?