Study guide
2.3 Service Enumeration & Operating System Detection
eJPT Junior Penetration Tester Exam
Track your progress as you read
Finished sections get a check mark.
1 Chapter 1: eJPT Foundations & Assessment Workflow
1.1 eJPT Exam Blueprint, Lab Architecture & Strategy 1.2 Penetration Testing Methodology & Standards
2 Chapter 2: Assessment Methodologies: Reconnaissance & Scanning
2.1 Open Source Intelligence (OSINT) & Information Gathering 2.2 Network Mapping & Port Scanning with Nmap 2.3 Service Enumeration & Operating System Detection
3 Chapter 3: Assessment Methodologies: Vulnerability Analysis
3.1 Service Vulnerability Identification & CVE Research 3.2 Vulnerability Impact Assessment & Prioritization
4 Chapter 4: Host & Network Auditing: System Enumeration
4.1 Linux Host & Network Enumeration 4.2 Windows Host & Privilege Enumeration 4.3 Network Services Auditing: SMB, FTP, SSH & Databases
5 Chapter 5: Host & Network Auditing: Data Collection & Credential Gathering
5.1 File Transfer Techniques for Linux & Windows 5.2 Credential & Password Hash Extraction
6 Chapter 6: Host & Network Penetration Testing: Exploitation & Metasploit
6.1 Exploit Analysis & Manual Modification 6.2 Metasploit Framework Architecture & Exploitation 6.3 Meterpreter Post-Exploitation Deep Dive
7 Chapter 7: Host & Network Penetration Testing: Pivoting & Password Attacks
7.1 Network Pivoting & Route Addition in Metasploit 7.2 Port Forwarding & SOCKS Proxy Tunneling 7.3 Network Service Brute-Forcing & Offline Hash Cracking
8 Chapter 8: Web Application Penetration Testing: Reconnaissance & Enumeration
8.1 HTTP Protocol Fundamentals & Burp Suite Interception 8.2 Directory & File Discovery with Gobuster
9 Chapter 9: Web Application Penetration Testing: Web Exploitation & Injections
9.1 Web Authentication Attacks & Login Brute-Forcing 9.2 SQL Injection (SQLi) Identification & Exploitation 9.3 Cross-Site Scripting (XSS), File Inclusion & Command Injection
2.3 Service Enumeration & Operating System Detection
Key Takeaways Explain This Topic AI
Service version detection ( -sV ) queries open ports with probe strings from nmap-service-probes and matches responses against compiled regular expressions to determine software names and version numbers.
Operating system fingerprinting ( -O ) analyzes TCP/IP stack implementation nuances, including initial Time-to-Live (TTL) values (Linux ~64, Windows ~128, Cisco ~255), TCP window sizes, and flag responses.
Manual banner grabbing with nc -nv or telnet provides rapid, direct verification of service software headers without relying on automated probing.
The Nmap Scripting Engine (NSE) automates deep service enumeration, authentication auditing, and vulnerability discovery using specialized Lua scripts organized into categories like default ( -sC ), vuln , and safe .
A standard penetration testing initial scan combines flags ( nmap -sC -sV -p- -T4 -oA target_scan <target> ) to maximize discovery while preserving structured artifacts.
Last updated: October 2026
Service Version Detection Mechanics (-sV)
Discovering that a TCP or UDP port is open provides only the initial step of vulnerability analysis. In enterprise networks and penetration testing environments, network services frequently run on non-standard ports (such as SSH running on port 2222 or HTTP running on port 8080). Conversely, standard ports may host custom or outdated daemons. To locate exploitable vulnerabilities and CVEs, a penetration tester must determine the exact service daemon and software version operating on each open port.
How Nmap Performs Service Detection ( -sV )
When invoked with the -sV flag, Nmap executes a multi-step probing sequence:
Initial Connection (NULL Probe) : Nmap establishes a TCP connection to the open port and listens silently for a specified timeout period. Many common daemons (including SSH, FTP, SMTP, and Telnet) automatically transmit an initial greeting banner upon connection establishment without waiting for client input.
Probe Progression via nmap-service-probes : If the service does not emit a greeting banner, Nmap consults its service probe database ( /usr/share/nmap/nmap-service-probes ). It transmits a sequence of standardized protocol challenge strings (such as HTTP GET requests, SSL/TLS Client Hellos, SMB negotiation requests, and RPC probes).
Signature Pattern Matching : When the service returns data, Nmap evaluates the response against a vast library of compiled regular expressions. These expressions extract:
- Product Name (e.g., Apache httpd , OpenSSH , vsftpd )
- Version Number (e.g., 2.4.41 , 8.2p1 , 2.3.4 )
- Operating System Platform (e.g., Ubuntu Linux , Windows )
- Hostname and Device Type (e.g., printer , embedded router )
- Controlling Probe Intensity ( --version-intensity )
Nmap categorizes each probe in nmap-service-probes with a rarity intensity rating ranging from 0 (very common) to 9 (rare). By default, Nmap operates at intensity level 7 :
--version-intensity <0-9> : Sets the probe depth. Higher values test more probes but increase scan duration.
--version-light : Shortcut for intensity level 2 . Fast scan that tests only high-probability probes.
--version-all : Shortcut for intensity level 9 . Tests every single probe against every open port, essential when identifying obscure services bound to non-standard high ports. # Scan designated ports with comprehensive version detection
- nmap -sV --version-all -p 80,443,8080,8443 target_ip
- Manual Banner Grabbing Techniques
While automated version detection is powerful, manual banner grabbing remains an essential skill for junior penetration testers. Automated tools can produce inaccurate deductions or trigger defenses, whereas manual socket interaction provides raw, ground-truth verification of service responses.
Banner Grabbing with Netcat ( nc )
Netcat ( nc ) is the premier utility for reading and writing raw data across TCP and UDP network connections: # Banner grab FTP (TCP 21)
- nc -nv 10.10.10.20 21
- # Expected Output: 220 (vsFTPd 3.0.3)
- # Banner grab SSH (TCP 22)
- nc -nv 10.10.10.20 22
- # Expected Output: SSH-2.0-OpenSSH_7.9p1 Debian-10+deb10u2
- # Banner grab SMTP (TCP 25)
- nc -nv 10.10.10.20 25
- # Expected Output: 220 mail.inlanefreight.local ESMTP Postfix (Ubuntu)
- Banner Grabbing Web Services with curl
For HTTP and HTTPS web servers, raw socket probes can be tedious due to HTTP request formatting requirements. The curl command retrieves HTTP response headers cleanly: # Fetch only HTTP headers using curl
curl -I http://10.10.10.20
- # Inspect response headers on secure web endpoints
- curl -k -I https://10.10.10.20
The returned Server: header (e.g., Server: Apache/2.4.29 (Ubuntu) ) and X-Powered-By: header (e.g., X-Powered-By: PHP/7.2.24 ) provide direct insight into the server stack.
Warning
Service banners can be forged or intentionally sanitized by security-conscious system administrators (e.g., setting ServerTokens Prod in Apache or using banner-masking modules). Always correlate banner data with underlying OS stack behavior and secondary enumeration artifacts.
Operating System Fingerprinting Mechanics (-O)
Determining the remote host's operating system is crucial for selecting compatible exploit payloads (such as differentiating between Windows x64 and Linux x86_64 architecture). Nmap implements active TCP/IP stack fingerprinting ( -O ), which evaluates how a target's networking stack handles subtle protocol edge cases.
TCP/IP Stack Fingerprinting Principles
While RFC standards govern TCP/IP networking, operating system vendors implement these standards with subtle internal variances. Nmap transmits a sequence of up to 16 distinct probe packets (including TCP SYN packets to open ports with varied TCP options, TCP packets with invalid flag combinations like SYN+FIN , TCP packets to closed ports, and ICMP Echo requests).
Nmap inspects the returned responses against its nmap-os-db signature database, analyzing key parameters:
Time-to-Live (TTL) : The initial TTL header value set by the sender operating system before packet transmission.
TCP Window Size : The initial receive window buffer negotiated during the handshake.
Don't Fragment (DF) Bit : Whether the operating system enforces IP path MTU discovery.
TCP Options Ordering : The precise sequence and formatting of TCP options (such as MSS, Window Scale, SACK Permitted, and Timestamps).
Response to Invalid Flags : How the stack responds to non-standard flag permutations (e.g., Linux returns RST/ACK to a SYN+FIN probe, while older Windows stacks ignore it or behave differently).
Analyzing Default TTL Values
Penetration testers can quickly infer a target operating system family simply by observing the TTL (Time-to-Live) value in received packets (via ping or tcpdump ):
- Operating System Family Initial Default TTL Observed TTL (Local Subnet) Observed TTL (2 Hops Away)
- Linux / Unix / Android / macOS 64 64 62 (64 - 2)
- Microsoft Windows 128 128 126 (128 - 2)
- Cisco IOS / Network Appliances 255 255 253 (255 - 2)
- Solaris / AIX 255 255 253 (255 - 2)
Because routers decrement the IP TTL header by 1 at each hop, a tester observing an inbound ICMP Echo Reply with ttl=63 can immediately deduce that the host is a Linux/Unix system operating one routing hop away (64 - 1 = 63).
The Nmap Scripting Engine (NSE)
The Nmap Scripting Engine (NSE) is an extensible subsystem powered by an embedded Lua interpreter. Located on Linux systems in /usr/share/nmap/scripts/ , NSE enables penetration testers to automate advanced service discovery, vulnerability detection, and credential auditing.
NSE Script Categories
Every NSE script belongs to one or more predefined categories:
default : Thoroughly tested, fast, reliable, and safe scripts executed automatically when the -sC flag is invoked.
vuln : Checks target services for specific known vulnerabilities, misconfigurations, and CVEs (e.g., checking for EternalBlue MS17-010 or Shellshock).
safe : Scripts that are non-intrusive and highly unlikely to crash target services, exhaust memory, or consume excessive bandwidth.
discovery : Queries public directories, SNMP trees, SMB shares, and DNS services to extract rich environmental metadata.
auth : Tests authentication credentials, audits default administrative passwords, and detects anonymous login access.
exploit : Actively attempts to deliver a functional exploit payload against the target service (use with extreme caution).
intrusive : Scripts that carry a significant risk of crashing target services or generating high network noise that triggers IDS alerts.
- Executing NSE Scripts # Run default scripts against open ports (equivalent to --script=default)
- nmap -sC target_ip
- # Run all scripts within the 'safe' and 'discovery' categories
- nmap --script "safe and discovery" target_ip
- # Target specific services with dedicated scripts
- nmap --script smb-os-discovery,smb-enum-shares -p 445 target_ip
- nmap --script ftp-anon,ftp-syst -p 21 target_ip
- nmap --script http-enum,http-title -p 80,443 target_ip
- # Execute a script with custom arguments using --script-args
- nmap --script smb-enum-shares --script-args smbuser=guest,smbpass="" -p 445 target_ip
- # Display documentation and arguments for a specific script
- nmap --script-help smb-os-discovery
- Assembling Effective Scan Strategies
In hands-on penetration testing examinations such as the eJPT, running an overly aggressive or unoptimized scan can exhaust lab bandwidth, cause target services to hang, or waste valuable examination time. Professional penetration testers utilize a disciplined, two-stage scanning methodology:
Stage 1: Rapid Full-Port Discovery
Quickly sweep all 65,535 TCP ports at an elevated rate to discover every active listening port without the overhead of service versioning or script execution: nmap -p- --min-rate 1000 -T4 -Pn -oN ports_quick.nmap target_ip
Stage 2: Deep Targeted Enumeration
Extract the discovered open ports and target only those active ports with service version detection ( -sV ), operating system fingerprinting ( -O ), default NSE scripts ( -sC ), and full output logging ( -oA ): # Example targeting discovered ports 21, 22, 80, and 445
nmap -sC -sV -O -p 21,22,80,445 -T4 -oA target_deep_scan target_ip
For smaller standalone lab hosts where time permits, combining flags into a single comprehensive command provides complete initial intelligence: nmap -sC -sV -p- -T4 -Pn -oA target_comprehensive target_ip
Common Network Services & Enumeration Cheat Sheet
Service Daemon Default Ports Transport Typical Service Banner Primary Manual Enumeration Primary Nmap NSE Script
- FTP 21 TCP 220 (vsFTPd 3.0.3) nc -nv <ip> 21 --script ftp-anon,ftp-syst
- SSH 22 TCP SSH-2.0-OpenSSH_8.2p1 nc -nv <ip> 22 --script ssh2-enum-algos
- Telnet 23 TCP Welcome to Linux telnet <ip> 23 --script telnet-encryption
- SMTP 25, 587 TCP 220 mail.domain.com ESMTP nc -nv <ip> 25 --script smtp-commands,smtp-enum-users
- DNS 53 UDP / TCP BIND version string dig @<ip> version.bind txt chaos --script dns-zone-transfer,dns-recursion
- HTTP / HTTPS 80, 443 TCP Server: Apache/2.4.41 curl -I http://<ip> --script http-enum,http-title
SMB / NetBIOS 139, 445 TCP Windows / Samba version enum4linux -a <ip> --script smb-os-discovery,smb-enum-shares
- SNMP 161 UDP MIB sysDescr string snmpwalk -v2c -c public <ip> --script snmp-info,snmp-sysdescr
- MySQL 3306 TCP 5.7.33-0ubuntu0.18.04.1 nc -nv <ip> 3306 --script mysql-info,mysql-enum
- RDP 3389 TCP CredSSP / TLS Cert rdesktop <ip> --script rdp-enum-encryption