WAF Là Gì? Web Application Firewall Bảo Vệ Ứng Dụng Web

What Is a WAF and Why Do You Need One?

A WAF (Web Application Firewall) is a security component that operates at Layer 7 of the OSI model, monitoring and filtering HTTP/HTTPS traffic between the internet and your web application.

Unlike traditional firewalls that only inspect IP addresses and ports, a WAF understands the semantics of HTTP requests — analyzing URLs, headers, cookies, query strings, and request bodies to detect attack patterns.

Why do you need a WAF?

Web applications are the number one attack target. According to the Verizon DBIR 2024 report, more than 40% of data breaches involve web applications. The OWASP Top 10 — the list of the ten most critical web security risks — can all be detected and blocked by a properly configured WAF.

WAF vs. Firewall vs. IDS/IPS:

How WAF Detection Works

1. Signature-based detection

Compares each request against a database of known attack signatures. Fast with low false positives, but only catches known attacks.

Request: GET /users?id=1' OR '1'='1
WAF rule: matches SQL injection pattern → BLOCK

2. Anomaly-based detection (Rule scoring)

Each request receives an anomaly score. Exceeding the threshold triggers a block. Less susceptible to bypass than pure signature matching.

3. ML-based / Behavioral detection

Learns a baseline of normal traffic and flags deviations. Effective against zero-days and unknown attacks, but more complex to tune.

OWASP Core Rule Set (CRS) is the most widely used open rule set for WAFs, covering:

• SQL Injection (REQUEST-942)
• XSS (REQUEST-941)
• Local File Inclusion (REQUEST-930)
• Remote Code Execution (REQUEST-932)
• Scanner detection (REQUEST-913)

ModSecurity and OWASP CRS in Practice

ModSecurity is an open-source WAF engine embedded into NGINX or Apache. Combined with OWASP CRS, it is the most popular self-hosted WAF solution.

Installing ModSecurity with NGINX:

1# Cài modsecurity và nginx connector
2apt-get install libmodsecurity3 libmodsecurity-dev
3git clone --depth 1 https://github.com/SpiderLabs/ModSecurity-nginx
4# Build nginx với modsecurity module

NGINX configuration:

1load_module modules/ngx_http_modsecurity_module.so;
2
3server {
4    listen 443 ssl;
5    modsecurity on;
6    modsecurity_rules_file /etc/nginx/modsec/main.conf;
7}

Example ModSecurity rule — detecting SQL injection:

SecRule ARGS "@detectSQLi" \
    "id:942100,\
    phase:2,\
    block,\
    capture,\
    t:none,t:utf8toUnicode,t:urlDecodeUni,t:removeNulls,t:removeComments,t:compressWhitespace,\
    msg:'SQL Injection Attack Detected via libinjection',\
    logdata:'Matched Data: %{TX.0} found within %{MATCHED_VAR_NAME}: %{MATCHED_VAR}',\
    tag:'application-multi',\
    tag:'language-multi',\
    tag:'platform-multi',\
    tag:'attack-sqli',\
    tag:'OWASP_CRS',\
    tag:'capec/1000/152/248/66',\
    tag:'PCI/6.5.2',\
    ver:'OWASP_CRS/3.3.4',\
    severity:'CRITICAL',\
    setvar:'tx.sql_injection_score=+%{tx.critical_anomaly_score}',\
    setvar:'tx.anomaly_score_pl1=+%{tx.critical_anomaly_score}'"

Each rule includes: id, phase (processing stage), block/detect, a transform chain, message, tags, severity, and anomaly score actions.

Popular WAF Solutions on the Market

AWS WAF:

• Managed service, integrates with CloudFront, ALB, and API Gateway
• Ready-made managed rule groups (AWS Managed Rules, Bot Control, Fraud Control)
• Manage rules via the Console or Terraform

 1resource "aws_wafv2_web_acl" "example" {
 2  name  = "example-waf"
 3  scope = "REGIONAL"
 4
 5  default_action { allow {} }
 6
 7  rule {
 8    name     = "AWSManagedRulesCommonRuleSet"
 9    priority = 1
10    override_action { none {} }
11    statement {
12      managed_rule_group_statement {
13        name        = "AWSManagedRulesCommonRuleSet"
14        vendor_name = "AWS"
15      }
16    }
17    visibility_config {
18      cloudwatch_metrics_enabled = true
19      metric_name                = "CommonRuleSetMetric"
20      sampled_requests_enabled   = true
21    }
22  }
23
24  visibility_config {
25    cloudwatch_metrics_enabled = true
26    metric_name                = "ExampleWAF"
27    sampled_requests_enabled   = true
28  }
29}

Cloudflare WAF: Managed and custom rules, easy to use, integrated with CDN and DDoS protection.

ModSecurity (self-hosted): Open-source, full control, low cost but requires self-maintenance.

WAF Bypass Techniques (Educational)

Understanding bypass techniques helps you tune your WAF more effectively and recognize its limitations.

URL Encoding bypass:

# Payload gốc (bị block):
' OR '1'='1

# URL encoded (có thể bypass WAF không decode đúng):
%27%20OR%20%271%27%3D%271

Case variation:

1# Original (blocked):
2SELECT * FROM users
3
4# Bypass attempt:
5SeLeCt * FrOm UsErS

Modern WAFs counter bypass attempts by: normalizing input before matching (lowercase, URL decode, multiple rounds of HTML decode), using parser-based detection instead of simple regex, and applying ML to detect evasion patterns.

Note: Knowledge of bypass techniques should only be used to test your own WAF (penetration testing) or to better understand how to configure WAF rules.

WAF Deployment Modes

Inline (Blocking mode):

• The WAF sits directly in the request path
• Bad requests are blocked immediately
• Risk: misconfiguration can block legitimate users
• Best suited after thorough rule tuning

Monitor (Detection-only mode):

• The WAF logs but does not block
• Ideal when first deploying, allowing you to observe false positives
• After 2–4 weeks of monitoring, tune rules, then switch to blocking

Best practices when onboarding a WAF:

1. Deploy in monitor mode
2. Run for 2–4 weeks, reviewing logs daily
3. Whitelist false positives (custom exclude rules)
4. Enable blocking mode for verified rules
5. Continue monitoring and tuning regularly

What is SQL Injection? How WAF blocks it

What is XSS? Cross-site scripting attack and defense

What is an API Gateway? Architecture and WAF integration

What is TLS 1.3? Advanced HTTPS encryption