Thursday, June 25, 2026

Cohort 3 | Project CloudIgnite Topics: Cost Optimization, AWS Trusted Advisor, AWS Support Plans, Whitepapers & Documentation, Knowledge Checks (Cost Management, Cloud Computing, Global Infrastructure), Lab 188 Tagging Resources, Lab 189 Cost Optimization & Pricing Calculator Duration: ~3 hours

---

Key Takeaways

• AWS Budgets sets spend/usage thresholds with alerts (including forecasted alerts when predicted to exceed); it only notifies, does NOT stop resources
• Cost-optimization techniques: right-size instances (match type to need, prefer T2/T3), use Lambda for spiky/low-volume workloads, use managed databases (RDS/DynamoDB), eliminate idle resources via CloudWatch metrics, Stopinator script for scheduled start/stop
• Trusted Advisor checks five pillars (Cost Optimization, Performance, Security, Fault Tolerance, Service Limits); core checks are free, full checks require Business or Enterprise support
• Support plan key facts: Basic = no technical cases; Developer = business-hours email, one contact; Business = 24/7 phone/chat, unlimited contacts, full Trusted Advisor; Enterprise = dedicated TAM, <15-min critical response
• Cloud benefits over on-premises: economies of scale → lower prices, stop guessing capacity, no racking/stacking/powering, increased speed & agility, go global in minutes (NOT: high latency or long procurement cycles)
• Global infrastructure: CloudFront uses Edge Locations for low latency; Region = geographic area with 2+ AZs; AZ = isolated fault domain with 1+ data centers; AZs linked by low-latency private connections
• Tags are case-sensitive key/value labels (up to 50 per resource) for cost allocation, automation, and organization; enforced via AWS Config (compliance) and IAM (creation-time requirement)

---

Table of Contents

1. Cost Optimization Strategies
2. AWS Trusted Advisor
3. AWS Support Plans
4. AWS Whitepapers & Documentation
5. Knowledge Check — Cost Management
6. Knowledge Check — Cloud Computing (Exam Prep)
7. Knowledge Check — Global Infrastructure (Exam Prep)
8. Lab 188 — Tagging AWS Resources (CLI)
9. Lab 189 — Cost Optimization & Pricing Calculator
10. CLF-C02 Exam Relevance — Consolidated Map
11. Glossary
12. Checkpoint Q&A Recap
13. Action Items & Housekeeping

---

1. Cost Optimization Strategies

This continued the Cost Management topic from June 24 (picking up around slide 12). The day was mostly about how to spend less on AWS.

Quick recap: AWS Budgets

• Lets you set a budget (e.g. $10/month) and get notified when a threshold of that budget is reached (e.g. 50%, 70%, 80%).
• Does NOT shut resources down — it only notifies you.
• Supports forecasted-cost alerts: if AWS predicts your month-end cost will exceed the budget, you get alerted early even if actual spend is still low.

Core cost-optimization techniques

📊 Visual: Cost-optimization techniques

The main levers for spending less on AWS — right-size, go serverless for spiky loads, use managed databases, hunt idle resources, and schedule start/stop.

[!NOTE] Free tier reminder from the instructor: a t3.micro with EBS + other services runs roughly $10–$13/month once outside the free tier. Idle-but-running resources still cost money after free tier ends.

[!TIP] EC2 vs Lambda cost logic — EC2 = pay for uptime (24/7). Lambda = pay per execution time. Bursty/low-traffic workloads are almost always cheaper on Lambda.

🎯 CLF-C02 Relevant

• Cost-optimization pillar of the Well-Architected Framework: right-sizing, serverless (Lambda) for variable demand, managed services, eliminating idle resources.
• Knowing which billing/cost tool does what (Budgets = alerts only, Cost Explorer = visualize, CloudWatch = metrics/idle detection) is a common exam theme.

---

2. AWS Trusted Advisor

Trusted Advisor inspects your account and gives recommendations across five pillars:

1. Cost Optimization (e.g. idle load balancers, unused Elastic IPs, underused instances)
2. Performance
3. Security
4. Fault Tolerance
5. Service Limits

• Free tier includes a limited set of core checks (the instructor referenced ~6 core checks); full/advanced checks require a Business or Enterprise Support plan.
• Uses the familiar traffic-light status: Green = no action needed · Yellow = worth checking / possible optimization · Red = critical, take immediate action.

📊 Visual: Trusted Advisor at a glance

Trusted Advisor checks your account across five pillars and reports a traffic-light status; only core checks are free, full checks need Business/Enterprise support.

🎯 CLF-C02 Relevant

• High. Trusted Advisor and its five check categories are frequently tested. Remember: full checks need Business/Enterprise support.

---

3. AWS Support Plans

AWS support is tiered. Higher tiers add faster response SLAs, more contacts, and a Technical Account Manager (TAM).

Response-time tiers (by severity)

[!WARNING] Basic Support has NO technical case support — only documentation, whitepapers, forums, the Personal Health Dashboard, and the 6 core Trusted Advisor checks.

📊 Visual: Which support plan?

Pick the tier by what you need — free docs/forums (Basic), business-hours email (Developer), 24/7 support + full Trusted Advisor (Business), or a dedicated TAM and fastest SLA (Enterprise).

🎯 CLF-C02 Relevant

• Very high. Support-plan comparison is a classic exam topic. Key facts to memorize:
  • TAM (Technical Account Manager) = Enterprise only.
  • < 15-min critical response = Enterprise only.
  • 24/7 phone/chat starts at Business.
  • Full Trusted Advisor requires Business or Enterprise.

---

4. AWS Whitepapers & Documentation

• Whitepapers are collections of technical documents covering topics like architecting best practices, security best practices, cloud economics, and serverless architecture.
• Available to everyone (including Basic) — a free knowledge/documentation resource.

🎯 CLF-C02 Relevant

• Knowing that whitepapers/documentation are free self-service resources (not gated behind paid support) is useful for support/resource questions.

---

5. Knowledge Check — Cost Management

Key Q&A from the cost-management KC:

[!NOTE] Watch for the classic mix-up: AWS Config enforces tags/compliance, while SCP controls permissions/access across accounts and OUs.

🎯 CLF-C02 Relevant

• High. Consolidated billing (Organizations), Config (tag enforcement), Budgets (spend alerts), and SCPs (access guardrails) are all core exam services.

---

6. Knowledge Check — Cloud Computing (Exam Prep)

The instructor noted these exam-prep questions closely mirror the real exam. Highlights:

• Pricing model = customers pay for resources on an as-needed basis (pay-as-you-go).
• Benefits of cloud over on-premises: massive economies of scale, stop guessing capacity, no racking/stacking/powering servers, increased speed & agility, go global in minutes.
• NOT benefits (trick answers): high latency, and multiple/long procurement cycles.
• Economies of scale = hundreds of thousands of customers aggregated in the cloud → AWS buys in bulk → passes savings on as lower prices.
• True/False: "You own the network-connected hardware" → False (AWS owns/manages the hardware).
• Cloud provides speed/agility, fault tolerance, disposable & temporary resources, and high availability.

📊 Visual: Cloud benefits vs distractors

The real benefits of cloud over on-premises — and the two classic trick answers (high latency, long procurement cycles) that are NOT benefits.

🎯 CLF-C02 Relevant

• Very high — this maps almost 1:1 to the exam's Cloud Concepts domain (benefits of cloud, economies of scale, pay-as-you-go, agility).

---

7. Knowledge Check — Global Infrastructure (Exam Prep)

📊 Visual: Global infrastructure hierarchy

How the pieces nest — a Region holds 2+ isolated AZs (each one or more data centers) linked by low-latency private links, while Edge Locations serve CloudFront closer to users.

[!TIP] Exam mnemonic: Edge Location → CloudFront/latency, Region → geography (2+ AZs), AZ → isolated fault domain (1+ data centers).

🎯 CLF-C02 Relevant

• Very high — Regions, AZs, and Edge Locations are guaranteed exam material in the Technology / Global Infrastructure domain.

---

8. Lab 188 — Tagging AWS Resources (CLI)

Goal: Use tags + the AWS CLI to filter, query, and bulk-manage (retag / stop / start / terminate) EC2 instances. (~45 min; steps 21–56, earlier steps pre-done.)

Connecting

• Windows → PowerShell, Mac/Linux → Terminal (same as prior labs).
• Delete your previous .pem file first, then download the fresh one — otherwise the connection fails (Permission denied = wrong/old key).
• Connect to the command host using its public IP (many instances have only private IPs; pick the one with a public IP).

Filtering & querying with the CLI

• Filter by tag:

  aws ec2 describe-instances --filters "Name=tag:project,Values=ERP system"
  

• Query specific fields (trim noisy output) with --query, e.g. return only Instance ID, Instance Type, Subnet ID, Image ID, AZ:

  aws ec2 describe-instances --query "Reservations[].Instances[].[InstanceId,InstanceType,SubnetId]"
  

• Tag filters must use the Name=tag:<key>,Values=<value> format — Name=project alone won't work.

Bulk tag / lifecycle scripts

• A bash script re-tagged resources (added a version tag, e.g. bumped to 1.1) only for instances tagged development.
• Stopinator script (written in PHP, uses the AWS SDK for PHP): stops/starts EC2 instances matching a set of tags.
  • Example: stop instances where project = ERP system AND environment = development. It scans all regions, then acts only on matching instances.
  • Add -s to start instead of stop. Instances without the target tags (e.g. the command host) are left untouched.
• Terminate script (terminate...php): terminates all EC2 instances that do NOT have an environment tag within a specific subnet (you supply the private subnet ID).

📊 Visual: Lab 188 — tag-driven CLI management

Connect, filter instances by tag, trim output with --query, then bulk-manage by tag — re-tag, Stopinator start/stop, or terminate untagged instances.

[!WARNING] The lifecycle scripts act on instances based purely on their tags — an untagged or mis-tagged production box could be wrongly stopped/terminated. Tag discipline directly affects safety here.

[!NOTE] Gotcha from the session: a missing semicolon in the PHP command caused "no EC2 instance found" until it was corrected. Instructor noted Python would be far easier than the PHP syntax used.

🎯 CLF-C02 Relevant

• Medium. The concept of tags for cost tracking, automation, and bulk resource management is exam-relevant. The exact CLI/PHP mechanics are hands-on skill, not exam content.

---

9. Lab 189 — Cost Optimization & Pricing Calculator

Goal: Downsize/right-size resources via CLI and estimate/compare costs with the AWS Pricing Calculator. (Nominally 50 min, mostly exploration.)

Setup — two instances at once

• Connect to both the CLI host and the cafe instance using two separate PowerShell windows (reuse the same PEM key — just change the IP in the connect command).
• Confirm you're on the right host: the prompt shows ec2-user@CLI-host vs the cafe/web server. If it shows your Windows path (user\Downloads), you're not connected.

Steps

1. From cafe instance: stop and remove MariaDB (stop then remove the server).
2. From CLI host: describe-instances to get the instance ID.
3. Stop the instance by ID, then change its instance type to downsize (e.g. T3 small → T3 micro) — smaller type = lower cost.
4. Explore Cost Explorer and the Pricing Calculator.

AWS Pricing Calculator

• No login required → Create estimate → pick Region → add services.
• Configured EC2 (shared, Linux, chosen instance type, EBS storage) and RDS (MariaDB, 20 GB, db.t3.micro), then viewed the summary (yearly cost).
• Optimization demo: dropping the EC2 type (T3 small → T3 micro) and EBS (40 GB → 20 GB) cut the yearly estimate by ~$100.

[!NOTE] RDS storage minimum is 20 GB — you can't go lower. Also, the live AWS UI is newer than the lab instructions, so some steps look outdated.

📊 Visual: Lab 189 — right-size & estimate

Remove the unused DB, downsize the instance type, then use the Pricing Calculator to compare estimates — dropping the type and EBS size cut ~$100/year.

🎯 CLF-C02 Relevant

• High. The AWS Pricing Calculator (estimate costs before deploying) and right-sizing to reduce cost are directly testable cost-management concepts. RDS vs self-managed DB and EBS sizing also appear.

---

CLF-C02 Exam Relevance — Consolidated Map

---

Glossary

• AWS Budgets — Set cost/usage budgets and receive alerts (including forecasted alerts). Notifies only; does not stop resources.
• Cost Explorer — Visualize and analyze cost/usage over time.
• AWS Pricing Calculator — Estimate the cost of a proposed architecture before deploying.
• Trusted Advisor — Automated best-practice checks across cost, performance, security, fault tolerance, and service limits.
• Technical Account Manager (TAM) — Dedicated AWS contact; Enterprise Support only.
• AWS Organizations — Central management of multiple accounts with consolidated billing.
• Service Control Policy (SCP) — Organization-level guardrail that allows/denies actions across accounts and OUs.
• AWS Config — Tracks resource configuration/compliance; can enforce required tags.
• Right-sizing — Matching instance type/size to actual workload demand to avoid waste.
• Stopinator — Script that starts/stops EC2 instances (by schedule or tag) to cut idle costs.
• Edge Location — CloudFront cache point close to users for low-latency delivery.
• Availability Zone (AZ) — One or more isolated data centers within a Region.
• Region — A geographic area containing two or more AZs.
• Economies of scale — Aggregated cloud demand lets AWS lower prices as usage grows.

---

Checkpoint Q&A Recap

1. Does AWS Budgets stop your resources when exceeded? → No. It only notifies (and can send forecasted alerts).
2. Best compute choice for 5 low-duration requests/day? → Lambda (pay per execution) rather than an always-on EC2 instance.
3. Which tool recommends across cost, performance, security, fault tolerance, and service limits? → Trusted Advisor.
4. Which support plan includes a dedicated TAM and < 15-min critical response? → Enterprise.
5. Which service gives consolidated billing across accounts? → AWS Organizations.
6. Which service enforces required tags? → AWS Config. Which controls access across accounts/OUs? → SCP.
7. What does CloudFront use for low-latency delivery? → Edge Locations.
8. Region vs AZ? → Region = geographic area with 2+ AZs; AZ = 1+ isolated data centers.
9. Is "high latency" a benefit of cloud? → No — it's a distractor. Real benefits: economies of scale, agility, no hardware ownership.

---

Action Items & Housekeeping

• Submit AND end each lab (Lab 188, Lab 189) before moving on.
• Work through the remaining KC cases over the next days (~10 questions / ~10 min each); a shared link with all case answers is available for review.
• Try KC cases in the Exam Preparation section — these mirror the real exam.
• Continue right-sizing practice in the Pricing Calculator (compare T3 small vs micro, EBS 40 vs 20 GB).
• Schedule: No class Saturday and none the following day (instructor at the AWS office) — next class is Monday.
• Progress note: the big/"chaotic" labs are mostly done; ~4–5 smaller labs remain. At-risk dashboard showed ~1,040 students green and ~63 yellow.

---