Network security lab demonstrating plaintext credential exposure in HTTP vs encrypted traffic in HTTPS using live Wireshark packet capture and analysis.
This lab validates a core SOC assumption: unencrypted protocols are a critical detection and data-theft vector.
This lab answers a fundamental network security question:
"What data is exposed to a network-level attacker intercepting HTTP vs HTTPS traffic?"
To validate this, I captured and analyzed live network traffic across both protocols and documented:
- Credential exposure in plaintext protocols
- TLS encryption effectiveness
- Detection opportunities for SOC analysts
- Network-level attacker capabilities
┌─────────────┐ HTTP (port 80) ┌──────────────┐
│ Browser │ ──────────────────────► │ Web Server │
│ Client │ ◄────────────────────── │ │
└──────┬──────┘ Plaintext Response └──────────────┘
│
│ Wireshark captures all traffic
▼
┌─────────────────────────────┐
│ Wireshark Packet Capture │
│ - Protocol dissection │
│ - Credential extraction │
│ - Traffic analysis │
└─────────────────────────────┘
┌─────────────┐ HTTPS (port 443) ┌──────────────┐
│ Browser │ ──────────────────────► │ Web Server │
│ Client │ ◄────────────────────── │ TLS 1.3 │
└─────────────┘ Encrypted Response └──────────────┘
│
│ Wireshark captures encrypted blobs only
▼
┌─────────────────────────────┐
│ Wireshark Packet Capture │
│ - TLS handshake visible │
│ - Payload: encrypted │
│ - No credential exposure │
└─────────────────────────────┘
| Data Type | Exposed? | Risk |
|---|---|---|
| Usernames | ✅ YES | Critical |
| Passwords | ✅ YES | Critical |
| Session tokens | ✅ YES | Critical |
| Form data | ✅ YES | High |
| Cookie values | ✅ YES | High |
Wireshark capture shows:
- Full HTTP GET/POST requests visible
- Credentials readable in packet payload
- No decryption required by attacker
| Data Type | Exposed? | Risk |
|---|---|---|
| Usernames | ❌ NO | None |
| Passwords | ❌ NO | None |
| Session tokens | ❌ NO | None |
| Form data | ❌ NO | None |
| Cookie values | ❌ NO | None |
Wireshark capture shows:
- TLS ClientHello/ServerHello visible
- SNI (hostname) visible — metadata only
- Payload: encrypted, unreadable
- Certificate exchange visible
| Tactic | Technique | Relevance |
|---|---|---|
| Credential Access | T1040 (Network Sniffing) | HTTP credential capture |
| Collection | T1557 (Adversary-in-the-Middle) | HTTP interception |
| Defense Evasion | T1573 (Encrypted Channel) | HTTPS protection |
Alert: Unencrypted credential submission detected
Logic: HTTP POST containing password= or passwd= fields
Severity: HIGH
Action: Alert SOC + block/redirect to HTTPS
Alert: Suspicious promiscuous mode detected
Rule: Wireshark/tcpdump process on non-analyst host
Severity: MEDIUM
Action: Investigate host for insider threat
Alert: HTTPS to HTTP redirect detected
Logic: 301/302 redirect from https:// to http://
Severity: HIGH
Action: Investigate for MITM or misconfiguration
| Tool | Purpose |
|---|---|
| Wireshark | Packet capture and protocol analysis |
| HTTP server | Plaintext traffic generation |
| HTTPS server | Encrypted traffic generation |
| Browser | Client traffic generation |
Screenshots in screenshots/ folder:
- HTTP packet showing plaintext credentials
- HTTPS packet showing encrypted payload
- Wireshark filter syntax for each protocol
- Side-by-side protocol comparison
- HTTP transmits credentials in plaintext — visible to any network observer
- HTTPS encrypts all payload data — only metadata (SNI, IP, timing) is visible
- SOC analysts should alert on any authentication over HTTP
- Network segmentation + TLS enforcement are critical security controls
- TLS certificate validation analysis
- SSL stripping attack simulation
- HSTS bypass techniques
- Certificate pinning bypass analysis
- Network-level DLP detection rules
GitHub: github.com/MrBipinShrestha LinkedIn: linkedin.com/in/shresthabipin Location: Sydney, Australia