HLD Mastery
Level first-deploy — First Deploy
LATENCY
ERRORS
COST
Diagram
Client
End Users
Internet
Network
Text
Annotation
Compute
EC2
Compute
Lambda
Serverless
Auto Scaling
EC2 Fleet
Spot Fleet
Spot Instances
Network
ALB
Load Balancer
API Gateway
REST / HTTP
CloudFront
CDN
WAF
Security
Route 53
DNS Routing
Database
RDS
Relational DB
Aurora
High Perf DB
DynamoDB
NoSQL
RDS Replica
Read Replica
DAX
DynamoDB Cache
Cache / Queue
ElastiCache
Cache
SQS
Message Queue
S3
Object Storage
SNS
Pub/Sub
Kinesis
Streaming
S3 Glacier
Cold Archive

Drag components onto the canvas to build your architecture.

Objectives0/3
Connect EC2 → RDS (direct or via cache)
Keep uptime above 95%
Sustain 95%+ uptime for 90 seconds
Coach
Next move
0/1
Wire EC2 → RDS (direct or via cache)

Start here: Connect EC2 → RDS (direct or via cache). Treat this like starter code: make one small architecture change, run it, then read the next coach step.

Metrics
Avg Latency
Error Rate
Throughput
Monthly Cost
Incidents
No active incidents
Event Log
00:00Canvas ready — hit RUN to start
Level first-deploy
First Deploy
Get the app live. Don't let it fall over.
Scenario

You've just joined a seed-stage startup. It's day one.

The CTO hands you a laptop and says: "We need the app live by Friday. It's Tuesday."

The stack is simple: a web server talking to a database. Your job is to wire it up and keep it running as real traffic arrives.

Starter
Starter scaffold

A small starting point is already on the canvas. Complete the missing pieces, run the simulation, and follow the coach as the system responds.

Objective

Connect EC2 to RDS. Keep your architecture alive as traffic ramps from 100 → 800 req/s.

Connect EC2 → RDS (direct or via cache)
Keep uptime above 95%
Sustain 95%+ uptime for 90 seconds
Watch For
EC2 CPU> 70%

Your app is stressed. Latency climbs. Users start noticing.

EC2 CPU> 95%

Requests are failing. Users see 503 errors. Score is dropping.

RDS Connectionsnear max

Database queries queue up. Everything downstream slows down.

Win Condition
Uptime ≥ 95% sustained for 90 seconds
Hint

If EC2 CPU spikes above 70%, click the node and upgrade the instance type. Bigger instance = more capacity.

"A single server has a limit. Every system does. This is yours."

SCORE1000
UPTIME100%
TIME00:00
FIRST DEPLOY