1 / 54

Network Reconnaissance and Enumeration

Network Reconnaissance and Enumeration . Section 3. Outline – Passive Enumeration. ARP Network Sniffing . Outline – Active Enumeration. Ping Traceroute Port/Service Scanning Banner Grabbing OS Fingerprinting Firewalking SNMP Routing Daemons. Objective.

xenia
Télécharger la présentation

Network Reconnaissance and Enumeration

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Network Reconnaissance and Enumeration Section 3

  2. Outline – Passive Enumeration • ARP • Network Sniffing

  3. Outline – Active Enumeration • Ping • Traceroute • Port/Service Scanning • Banner Grabbing • OS Fingerprinting • Firewalking • SNMP • Routing Daemons

  4. Objective • Targeted Network Reconnaissance • Network topology • Host systems • Host applications/services • Host configuration • Host users accounts • Naming conventions

  5. Passive Enumeration • Generally LAN based • Generally undetectable

  6. ARP Enumeration • All systems with IP stacks will have ARP • Local ARP table contains a list of all MAC addresses mapped to IP addresses • First 6 bytes of MAC address identify card/system vendor (e.g) • 00067Cxxxxxx : Cisco • 080020xxxxxx : Sun • ARP mappings are gathered and released automatically by the system

  7. ARP Enumeration • Use the ‘arp -a’ command to get access to the local ARP table C:\>arp -a Interface: 158.177.248.190 on Interface 4 Internet Address Physical Address Type 158.177.248.1 00-00-0c-07-ac-02 dynamic 158.177.248.2 00-d0-d3-38-68-08 dynamic 158.177.248.3 00-d0-d3-36-08-84 dynamic 158.177.248.191 00-d0-59-05-d2-13 dynamic

  8. Network Sniffing • Placing a network adapter in ‘promiscuous mode’ enables the reception of all network traffic passing the adapter. • Most effective on non-switched networks • Even on switched networks all broadcast traffic may be received • No valid IP address required

  9. Network Sniffing – What do we get? • MAC addresses • IP addresses and IP structure • RIP, OSPF packets • System names • NetBIOS broadcasts • Cisco Discovery Protocol • IPX service announcements • Default gateways • Not to mention usernames, passwords and data…but more on that later

  10. Network Sniffing – The Tools • Windows • NetMon (Microsoft) • Sniffer Pro (NAI) • Tcpdump + winpcap (GPL) • UNIX • Tcpdump (GPL) • Etherreal (GPL) • Dsniff (more attack than recon – GPL)

  11. Network Sniffing - Filtering • On high throughput networks use capture filters • Whole packet is usually not needed – just headers and start of data portion • Any IP or TCP/UDP protocol may be included or excluded from a capture filter • ‘Noisy’ addresses may be filtered out • All network sniffing tools support filtering to some degree • For extended periods of sniffing filters will be essential

  12. Network Sniffing - tcpdump • Console based, highly configurable, low resource usage.

  13. Network Sniffing – Ethereal

  14. Active Enumeration • Can yield a great deal of data • Can be detected • May require specialised software/OS

  15. The Mighty Ping • Used to determine host reachability • Nearly all systems with an IP stack have a ping facility • Uses the ICMP protocol • Provides • Round trip time (RTT) • Time to live (TTL) – indicates hop-count • Generally unnoticed if used in moderation

  16. Ping test H A C K E R T A R G E T ICMP Echo Request ICMP Echo Response

  17. Ping - Usage • Ping is used on one system at a time • It is possible to ping a broadcast address, sometimes multiple responses are returned C:\>ping 158.177.248.1 Pinging with 32 bytes of data: Reply from : bytes=32 time=1ms TTL=255 Reply from : bytes=32 time=1ms TTL=255 Reply from : bytes=32 time=1ms TTL=255 Reply from : bytes=32 time<10ms TTL=255 Packets: Sent = 4, Received = 4, Lost = 0 (0% loss) RTT: Minimum = 0ms, Maximum = 1ms, Average = 0ms

  18. Ping – Sweeping • Ping-sweeping is a term for sweeping multiple addresses with a ping query to determine host reachability. • Used for finding targets within a network range. • Will not work on targets that have ICMP traffic filtered or blocked. • Tools such as NMAP, MingSweeper and Pinger will perform a ping-sweep.

  19. More with ping Identifying the Firewall Good commercial firewalls block time stamp requests and source routed packets. Use ping to send them, and it will show there is a firewall in the way # ping -v -T tsandaddr 10.0.0.120 PING 10.0.0.120 (10.0.0.120) from 10.0.0.1 : 56(124) bytes of data. --- 10.0.0.120 ping statistics --- 16 packets transmitted, 0 received, 100% loss # ping -v -T tsandaddr 10.0.0.125 PING 10.0.0.125 (10.0.0.125) from 10.0.0.1 : 56(124) bytes of data. --- 10.0.0.125 ping statistics --- 8 packets transmitted, 0 received, 100% loss

  20. More fun with ICMP Identifying the Firewall W Richard Stevens – a god amongst men and author of TCP/IP Illustrated etc left us with the ability to to generate ICMP address mask requests and ICMP time requests # icmpaddrmask 10.0.0.120 Received mask fffff00, from 10.0.0.120 #

  21. Traceroute • Used to determine both connectivity AND the IP route used to reach the target. • Available on most but not all systems that have an IP stack. • Relies on the fact that routers should always decrement the TTL of a packet as it passes. • Generally uses ICMP packets on Windows and UDP packets on UNIX systems. However any IP protocol could be used in theory.

  22. Traceroute - Visual route Probably the Firewall or the perimeter router

  23. Port/Service Scanning • Used to determine what TCP or UDP ports are available on a target system. • The scanner will attempt to connect to each port on the target. • The scanner should detect the port in one of three states: • Closed – port reachable but no service present • Open – port is reachable and service is present • Filtered – port is non-reachable, possible firewall or packet filter is present

  24. Background to TCP scanning • Determine what ports of a host are listening for connections • 4 main types of TCP scans • TCP connect() • SYN scan • FIN scan + Null + ACK + XMAS • Fragmentation scanning

  25. TCP Connect Scan • Uses a system call provided by the operating system • If the port is listening, connect will succeed • Does not require any special privileges • Easily detectable • Most reliable • Cannot usually detect filtered ports

  26. The Three-way handshake TCP Connect Syn Syn/Ack Ack C L I E N T T A R G E T

  27. SYN Scan • Commonly referred to a half open scanning • Sends a SYN packet and waits for a response • A SYN/ACK response indicates port is listening • A RST packet indicates the port is not listening • Less likely to be logged • No response indicates port is filtered • Requires raw sockets requiring root or Administrator privileges • Some IDS confuse this with a SYN flood

  28. FIN Scan + Null + XMAS • More stealthy than a SYN scan • Reply with a proper RST packet indicates that the port is closed • Open ports tend to ignore the FIN packet • However Microsoft boxes tend to send a RST packet regardless • Can be used to differentiate a Unix box from a Microsoft box

  29. ACK Scan • Not used to identify open and closed ports • Used to identify filtered ports • Can be useful for mapping packet filter rules

  30. Fragmentation scanning • A modification of other techniques of scanning • Breaks the probe packet into a couple of small IP fragments • Breaking up the TCP header into several smaller packets makes it harder to detect and some packet filters will pass the packet unchecked rather than wait for all of the fragments to arrive.

  31. UDP Port Scanning • Single UDP packet to each port being tested • Closed ports respond with an ICMP unreachable message. • Open ports will NOT respond • Filtered ports will NOT respond • Results can be ambiguous on filtered targets • Can be very slow due to ICMP message rate limiting which is specified within the RFC’s describing IP & ICMP. Some systems do not implement this (e.g. all Windows systems.)

  32. Banner Grabbing • The process of examining banner strings returned by services bound to open TCP ports • Enables identification of service applications including software version. • Not effective for non-character based services (e.g. SMB, HTTPS) • Some service applications will not send banner information until prompted (e.g. HTTP) • Useful for identification of service applications on non-standard ports

  33. Banner Grabbing – Manual Testing • Use netcat or telnet to connect to the port : C:\nc 10.0.0.120 25 220 Sendmail/8.8.8 ESMTP Looks like Sendmail 8.8.8 mail server - easy

  34. Banner Grabbing - Automation

  35. Fingerprinting • Stack fingerprinting is used to determine the operating system of a target host • Utilises differences in the implementation of the IP stacks • Involves sending non-standard packets to the target and examining any responses • Not always accurate • Very easy to spot for IDS

  36. Fingerprinting – A Simple Test • A single ping can be used to aid in OS detection and is a very basic way of fingerprinting a target. Windows !! # ping 10.0.0.120 PING 10.0.0.120 : 56(84) bytes of data. 64 bytes from 10.0.0.120: icmp_seq=1 ttl=128 time=0.280 ms --- 10.0.0.120 ping statistics --- • Note that TTL=128 in the reply. That almost guarantees that the target is a Windows system of some description.

  37. Fingerprinting – A Simple Test • Here are the default TTL (Time To Live) values for a few common systems : • Cisco Devices 255 • Most Windows Systems 128 • Windows 95 32 • Linux <= 2.0.x 64 • Linux >= 2.1.x 255 • Solaris 255

  38. Advanced IP Stack Fingerprinting • Involves sending crafted packets to the target • Ideally requires >=1 open port and >=1 closed port • Packet filters, firewalls and transparent proxies can render IP stack fingerprinting useless when using automated tools such as NMAP, Queso or MingSweeper because they sometimes re-write packets.

  39. Advanced IP Stack Fingerprinting • Tools for automated stack fingerprinting • NMAP stack fingerprinting • Xprobe ICMP stack fingerprinting • MingSweeper combined NMAP+ICMP • Queso – early stack fingerprinting, NMAP draws tests from this software.

  40. NMAP • “Network Mapper” • Open source utility for network exploration • Its functions include a wide variety of port scanning mechanisms, OS detection and ping sweeps. • Runs on most Unix based operating systems • Has a graphical user interface • Its FREE!

  41. # nmap -sS -n -p 1-10000 10.0.0.125 Starting nmap 3.48 ( http://www.insecure.org/nmap/ ) All 10000 scanned ports on 10.0.0.125 are: filtered Nmap run completed -- 1 IP address (1 host up) # nmap -sU -n -p 1-10000 10.0.0.125 Starting nmap 3.48 ( http://www.insecure.org/nmap/ ) All 10000 scanned ports on 10.0.0.125 are: filtered Nmap run completed -- 1 IP address (1 host up)

  42. Mingsweeper • Windows based network reconnaissance utility • Performs ping sweeps, Reverse DNS sweeps, TCP & UDP port scans, OS identification and application identification. • It is also FREE!

  43. Scanning - Ming-sweeper

  44. Stack Fingerprinting Test Probes • NMAP tests comprise of 9 probes • TClass TCP Initial Sequence Number probe • T1 to T4 Four TCP open port probe • T5 to T7 Three TCP closed port probe • PU Port Unreachable UDP probe • MingSweeper adds 4 probes • I1 Four ICMP echo probes • I2 Four ICMP timestamp probes • I3 Four ICMP address mask probes • I4 Four ICMP information query probes

  45. Firewalking • Attempting to route packets through firewall • Vary IP protocol, TCP/UDP port numbers • Use decrementing TTL values to illicit responses from devices behind firewall • Not particularly successful on modern firewall systems

  46. Firewalking • Hping – useful tool for generating many types of hand-crafted IP packets. UNIX only. • Mptraceroute – performs traceroutes using multiple IP protocols and allowing port specifications. Win 2000/XP only.

  47. Identifying the FW – Ike-scan # ike-scan -v 10.0.0.125 Starting ike-scan 1.6 with 1 hosts --- Pass 1 of 3 completed --- Pass 2 of 3 completed --- Pass 3 of 3 completed Ending ike-scan 1.6: 1 hosts scanned in 22.595 seconds (0.04 hosts/sec). 0 returned handshake; 0 returned notify

  48. Identifying the Firewall - LFT # lft -vv –E -n 10.0.0.120 Looks like we made it. Everyone responded. Moving on... Will finish TWO Concluding with 2 hops. TTL LFT trace to 10.0.0.120:80/tcp **[4.2 BSD bug]next gateway may errantly reply with reused TTLs 1 [target] 10.0.0.120:80 6.5ms **[4.2 BSD bug]next gateway may errantly reply with reused TTLs 2 [target] 10.0.0.120:80 1.6ms

  49. SNMP Enumeration • Community string is equivalent to a password used to control access to node information • Very often set to ‘public’ and ‘private’ • All transmissions are in clear text • SNMP daemons found on many network devices and hosts systems. • Generally all SNMP nodes in a network share the same community name

  50. SNMP Enumeration • SNMP community names can be brute forced easily • SNMP can yield a huge amount of information • ARP tables • Routing tables • Device specific information • Traffic statistics • Host based SNMP daemons may yield process, service and user information (e.g. Windows NT SNMP service)

More Related