Web Application Firewall (WAF)

WAF protects web applications from common exploits like SQL injection, cross-site scripting (XSS), and other OWASP Top 10 vulnerabilities. It’s the Layer 7 complement to Azure Firewall’s Layer 3/4 protection.

🎬 Video coming soon

Web Application Firewall (WAF)

Web Application Firewall (WAF)

~11:00

WAF on Application Gateway vs Front Door

WAF: Application Gateway vs Front Door

Feature	WAF on Application Gateway	WAF on Front Door
Scope	Regional (per Application Gateway)	Global (at edge PoPs worldwide)
Rule set	OWASP CRS 3.2, 3.1, 3.0, 2.2.9	Default Rule Set (DRS) 2.1, 2.0, 1.1
Custom rules	Yes	Yes
Managed rules	Yes (OWASP Core Rule Set)	Yes (DRS — Premium tier only)
Bot protection	Limited	Yes (Premium tier)
Geo-filtering	Via custom rules	Built-in
Rate limiting	Via custom rules	Built-in
Per-site policies	Yes (per listener)	Yes (per route)
Best for	Regional web apps, internal applications	Global web apps, CDN-fronted applications

Exam Tip: On Front Door, managed WAF rules (DRS) and bot protection require the Premium tier. Standard Front Door only gets custom WAF rules. On Application Gateway, WAF is available as a WAF v2 SKU (no separate tier needed).

Detection vs Prevention Mode

Mode	Behaviour	When to Use
Detection	Logs attacks but does NOT block them. Traffic passes through.	Initial deployment — learn what gets flagged.
Prevention	Logs AND blocks attacks. Malicious requests get a 403 Forbidden.	Production — after tuning false positives.

Best practice workflow:

1. Deploy in Detection mode — let WAF log for 1-2 weeks
2. Monitor logs — identify false positives (legitimate traffic flagged as attacks)
3. Create exclusions — for specific request headers, cookies, or body fields that trigger false positives
4. Switch to Prevention mode — WAF now blocks actual attacks while allowing legitimate traffic
5. Continuously monitor — new false positives may appear as application changes

Exam Tip — False Positive Management: The exam tests that you know the recommended workflow. Don’t deploy straight to Prevention mode — you’ll block legitimate users. Detection first, exclusions, then Prevention.

💡 WAF False Positive Management

False positives are the biggest operational challenge with WAF. Here is how to handle them without weakening your protection:

1. Per-field exclusions (preferred): Exclude specific request headers, cookies, or body fields from a specific rule. Example: exclude the “comment” form field from SQL injection rule 942130. This is the most targeted fix.

2. Custom allow rules: Create a custom rule to allow traffic from known-good sources (monitoring IPs, partner APIs) before managed rules evaluate it. Custom rules process first.

3. Per-rule disabling: Disable a specific managed rule ID that causes persistent false positives. Only disable the individual rule, never an entire rule group.

4. Never disable the entire WAF: If you have too many false positives, the fix is better exclusions and custom rules — not turning WAF off. Switching back to Detection mode temporarily while you tune is acceptable.

WAF Policies

A WAF policy is a standalone resource containing all WAF configuration:

Components of a WAF policy:

Component	Purpose
Managed rules	Pre-built rule sets from Microsoft (OWASP CRS or DRS). Enable/disable individual rule groups.
Custom rules	Your own rules based on IP, geo, rate limiting, or request attributes. Processed before managed rules.
Exclusions	Specific request fields to exclude from rule evaluation (fix false positives).
Mode	Detection or Prevention.
Policy settings	Request body inspection, file upload size limits, etc.

Custom rules are processed before managed rules. This means your custom Allow rules can whitelist traffic before managed rules would block it.

Associating WAF Policies

Platform	Association Level
Application Gateway	Per gateway (global), per listener (per-site), or per URL path map
Front Door	Per endpoint (global) or per route (per-site)

A more specific association overrides a broader one. A per-listener policy overrides the per-gateway policy for that listener’s traffic.

Custom Rules — Common Patterns

Block specific IPs or countries: Block requests from IP ranges known for abuse, or restrict access to specific countries for compliance.

Rate limiting: Limit the number of requests per client IP within a time window. Prevents brute-force and scraping attacks.

• Set threshold (e.g., 100 requests per minute per IP)
• Action: Block, Log, or custom response

Allow known-good traffic: Whitelist specific IPs (e.g., monitoring services, partner APIs) so managed rules don’t block them.

Azure Firewall vs WAF — They’re Complementary

Feature	Azure Firewall	WAF
Layer	Layer 3/4 (+ limited L7 with FQDN)	Layer 7 (HTTP/HTTPS only)
Scope	All traffic (TCP, UDP, ICMP, HTTP)	HTTP/HTTPS only
OWASP protection	No	Yes — core capability
SQL injection / XSS	No	Yes
FQDN filtering	Yes	Not the primary purpose
IDPS	Premium only	No (different attack types)
Deployment	Hub VNet / VWAN hub	On Application Gateway or Front Door
Relationship	Network-level firewall	Application-level firewall

They work together: Azure Firewall filters network traffic (block bad IPs, restrict outbound, FQDN filtering). WAF protects web applications from application-layer attacks (SQL injection, XSS, OWASP). Deploy both in enterprise architectures.

🔒 Aisha’s defence-in-depth:

Internet → Front Door WAF (global, blocks OWASP attacks)
  → Azure Firewall (hub, filters network traffic)
    → Application Gateway WAF (regional, additional rules)
      → Backend VMs

Key Takeaways

• WAF on Application Gateway: regional, OWASP CRS
• WAF on Front Door: global at edge, DRS (Premium only for managed rules)
• Always deploy Detection mode first, then Prevention after tuning
• Custom rules process before managed rules
• Azure Firewall and WAF are complementary — use both for defence-in-depth

---

Test Your Knowledge

Question

What is the recommended workflow for deploying WAF?

Click or press Enter to reveal answer

Answer

1. Deploy in Detection mode. 2. Monitor logs for 1-2 weeks. 3. Create exclusions for false positives. 4. Switch to Prevention mode. 5. Continuously monitor.

Click to flip back

Question

Which Front Door tier provides managed WAF rules (DRS) and bot protection?

Click or press Enter to reveal answer

Answer

Premium tier only. Standard Front Door provides custom WAF rules but not managed rule sets (DRS) or bot protection.

Click to flip back

Question

Are custom rules processed before or after managed rules?

Click or press Enter to reveal answer

Answer

Custom rules are processed BEFORE managed rules. This allows you to whitelist known-good traffic with custom Allow rules before managed rules would potentially block it.

Click to flip back

Question

How do Azure Firewall and WAF differ?

Click or press Enter to reveal answer

Answer

Azure Firewall: Layer 3/4 network firewall (all traffic types, FQDN filtering, IDPS in Premium). WAF: Layer 7 application firewall (HTTP/HTTPS only, OWASP protection, SQL injection, XSS). They're complementary — use both.

Click to flip back

---

Knowledge Check

Sam deploys WAF in Prevention mode immediately. Customers report they can't submit forms on his website. What went wrong?

AWAF is blocking all POST requestsBManaged rules flagged legitimate form data as SQL injection — false positivesCPrevention mode requires a separate licenceDWAF doesn't support form submissions

Check Answer

Knowledge Check

Aisha needs both network-level filtering and OWASP protection for her web applications. What should she deploy?

AAzure Firewall only — it handles all securityBWAF only — it covers all attack typesCAzure Firewall for network filtering AND WAF for application protectionDNSGs are sufficient for both

Check Answer

---

Congratulations! 🎉

You’ve completed all 26 modules of the AZ-700 study guide. You’ve covered:

• Domain 1: Virtual Networks, IP addressing, DNS, routing, peering, monitoring, and DDoS protection
• Domain 2: Site-to-Site VPN, Point-to-Site VPN, ExpressRoute, Virtual WAN, and hybrid connectivity design
• Domain 3: Load Balancer, Traffic Manager, Application Gateway, Front Door, and the load balancing decision matrix
• Domain 4: Private Link, private endpoints, DNS for private endpoints, and service endpoints
• Domain 5: NSGs, ASGs, flow logs, Azure Firewall, Firewall Manager, and WAF

What’s next:

• Review the flashcards and quizzes — aim for 100% on all quiz questions
• Practice with hands-on labs in your Azure subscription
• Take practice exams to identify weak areas
• Revisit modules where you scored lowest on practice exams

You’ve got this! 💪