Azure AZ-204 study plan
Azure AZ-204 · Give each week a clear focus.
8 weeks · About 6 hours per week
Develop Azure compute solutions
Create an App Service plan and deploy a web app into it using the CLI, then add a deployment slot and complete a slot swap
Configure app settings and connection strings as slot-sticky versus slot-swappable and verify behavior after a swap
Enable autoscaling on the App Service plan with a CPU-based rule and confirm scale-out by driving load
Turn on application logging and stream the log output while exercising an endpoint
Your checkpoint
A CLI script set in a repo that provisions the plan, app, two slots, and scale rules from scratch, plus a written post-swap verification checklist you actually ran through
Develop Azure compute solutions
Build a Docker image, push it to Azure Container Registry, and deploy it to Web App for Containers with a registry credential
Run the same image as an Azure Container Instance with environment variables and a non-default restart policy
Create a Function app with an HTTP and a timer trigger, and set the runtime and plan explicitly
Implement a durable function orchestration with a chained activity and inspect the orchestration status endpoint
Your checkpoint
A repository containing the Dockerfile, registry push and container deploy scripts, and a deployed Function app whose durable orchestration completes successfully when invoked
Develop Azure compute solutions & Develop for Azure storage
Wire a blob-triggered function to a queue output binding and verify the message lands in storage end to end
Add a poison-queue path to a queue-triggered function and prove a failing message is retried then parked
Externalize function configuration into App Configuration and consume it through Key Vault references
Enable Application Insights on the function app and follow a single invocation across the trigger, binding, and output
Your checkpoint
A deployed serverless pipeline (blob in, queue out, poison handling) with the binding configuration committed and one traced invocation exported as evidence
Develop for Azure storage
Move blobs across hot, cool, and archive tiers and apply a lifecycle management rule that you confirm by policy evaluation
Generate a service SAS and a stored access policy, then revoke access by editing the policy without rotating the key
Set and query blob index tags and metadata to filter a container listing
Create a Cosmos DB container with a partition key and compare two consistency levels under the same read pattern
Read the Cosmos DB change feed from a function and process inserted documents
Your checkpoint
A storage and Cosmos DB lab notebook (commands plus observed outputs) covering tiers, SAS revocation, tag queries, and a working change-feed processor
Implement Azure security
Register an app in Microsoft Entra ID and acquire a token with MSAL using both an interactive flow and a client-credentials flow
Enable a system-assigned managed identity on an App Service app and grant it least-privilege access to Key Vault
Read, write, and rotate a secret and a certificate from Key Vault using the SDK rather than the portal
Request a scoped Microsoft Graph permission and query the signed-in user's profile with the resulting token
Your checkpoint
A deployed app that reads its secret from Key Vault through managed identity with no credentials in code or configuration, plus the exact role assignment used and why
Implement Azure security & Monitor, troubleshoot, and optimize Azure solutions
Protect a web API with a bearer-token requirement and reject both an expired and a wrong-audience token
Write KQL queries against Log Analytics to surface failed requests, slow dependencies, and their correlation IDs
Instrument an app with custom events and metrics, then locate them in Application Insights after enabling sampling
Add a caching layer for a hot read path and measure the latency change before and after
Your checkpoint
A troubleshooting runbook with four saved KQL queries and a before/after latency table from the caching change
Monitor, troubleshoot, and optimize Azure solutions & Connect to and consume Azure services and third-party services
Call an unreliable downstream API through HttpClientFactory with a Polly retry and circuit-breaker policy and force each path
Publish an API through API Management and apply a rate-limit and a header-rewrite policy
Send and receive Service Bus messages with peek-lock, complete/abandon, and inspect the dead-letter queue
Publish an Event Grid custom topic event and subscribe a function handler to it
Diagnose a deliberately broken deployment using App Insights transaction traces and dependency failures
Your checkpoint
A resilience and messaging demo repo showing a retry, a circuit-breaker trip, a dead-lettered message, and the trace that identified the broken deployment
Connect to and consume Azure services and third-party services & Monitor, troubleshoot, and optimize Azure solutions & Develop Azure compute solutions
Stream events into Event Hubs with a partition key and read them back with a checkpointing consumer
Consume a third-party HTTP API with pagination and throttling handled correctly
Re-run your two weakest labs from weeks 1 through 7 without notes
Take a timed full-length practice exam and tag every miss by domain and by cause (knowledge gap versus misread)
Your checkpoint
A remediation list ranked by domain weight with every item closed out, plus a final practice-exam score recorded alongside your week-4 baseline
Milestones
Week 3: Deploy a blob-to-queue serverless pipeline from an empty subscription in under 45 minutes with no portal clicking
Week 4: Finish a 40-question practice set covering domains 1 and 2 at 70% or better, and confirm no domain-1 or domain-2 objective is still marked unknown
Week 6: Answer four KQL troubleshooting questions and two token-validation questions correctly without opening documentation
Week 8: Score 80% or better on a timed full-length practice exam with no domain below 70%