Skip to content
Tutorials Center
My learning

Topics

Lesson 15 of 38 ยท Scanning and Enumeration

Host Discovery: Finding Live Systems

Learn how testers find live hosts with ICMP, ARP, TCP and UDP probes, how firewalls hide hosts, and how to plan and log a discovery sweep within scope.

Recon told you which names and address ranges belong to the client. The scanning phase begins with a simple question: which of those addresses actually have a machine behind them? Host discovery (also called a ping sweep or network mapping) answers it. Scanning every port of every address would be slow and noisy, so you first reduce the list to live hosts, then go deeper on those. This lesson explains the probes, why they sometimes fail, and how to keep the sweep safe and in scope.

Scanning is active. Before sending anything, make sure every address is in your written scope (Rules of Engagement).

How a host reveals itself

A host that is up responds to certain packets. Different probes work in different situations.

Probe How it works When it is useful
ICMP echo (ping) Send echo request; a live host answers with echo reply Quick and simple; often blocked by firewalls
ARP request Ask "who has this IP?" on the local network Most reliable on the same subnet; cannot cross routers
TCP SYN to common ports Send SYN to 80 or 443; a SYN/ACK or RST means a host is there Works when ping is blocked
TCP ACK probe Send ACK; a live host replies RST Can reveal hosts behind simple stateless filters
UDP probe Send UDP to a likely service; ICMP port unreachable means it is up Slower; many hosts stay silent

A live host does not need to answer every probe. A firewall may drop pings yet allow web traffic, so a good sweep combines several probe types.

Probe types

ICMP echo Classic ping. Often filtered.
ARP Same subnet. Very reliable.
TCP probes SYN to ports 80 and 443.
UDP probes Slow but sometimes needed.

Combine probes to avoid missing quiet hosts.

1 / 2

Internal versus external discovery

On an internal network you can use ARP, which cannot be blocked by host firewalls because it is how Ethernet works. You see every device on the subnet, including printers, cameras and phones. On an external target, ARP is unavailable (it does not cross routers). You rely on ICMP and TCP probes, and the results depend on firewall rules: a quiet range does not mean empty.

You may also look at passive sources: the ARP table of a compromised or provided machine, DHCP leases, and network monitoring logs that the client shares in a grey box test.

Read-only: example nmap discovery

nmap is the standard network mapper. These commands are read-only here, with example output in the real format using documentation addresses.

Terminal
nmap -sn 192.0.2.0/28
TEXT
Starting Nmap 7.x ( https://nmap.org ) at 2026-05-12 09:30 UTC
Nmap scan report for 192.0.2.1
Host is up (0.0021s latency).
Nmap scan report for 192.0.2.10
Host is up (0.0040s latency).
Nmap scan report for 192.0.2.14
Host is up (0.0033s latency).
Nmap done: 16 IP addresses (3 hosts up) scanned in 1.8 seconds

-sn means "no port scan: host discovery only". To list targets without sending anything, use -sL; this is handy for checking that your ranges match the scope. To skip ping and treat every address as up, use -Pn, which is slower but needed when ICMP and common probes are blocked.

Rate, timing and safety

Aggressive timing can overwhelm weak devices or trigger alarms. Use polite settings first (for example -T2 or -T3) and increase only if the RoE permits. Log the start and end times and the command. Exclude fragile addresses with --exclude. If a host stops responding mid-scan, pause, tell the technical contact, and note it in the log.

False negatives and false positives

  • A host can be up but silent because every probe is filtered. Try different probes and ports, and use -Pn carefully.
  • A host can appear up when it is not: a firewall or load balancer may answer for many addresses (a "proxy ARP" or SYN proxy). If every address in a range replies, suspect it.
  • Hosts can be intermittently online: laptops sleep. Repeat the sweep at other times if needed.

Stand-in: simulate a ping sweep

The first example simulates probing a /28 network. A dictionary describes what each invented host answers; the sweep combines several probes so that a host that ignores ping can still be found.

Python Combine probes to find live hosts

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The second example estimates how long a sweep takes at different rates and checks that no excluded address is probed. This is the kind of arithmetic that goes into the plan.

Python Plan the sweep: time and exclusions

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Recording results

Write down what you found, how, and when. A table with the address, the probe that answered, the time, and a note on any oddity is enough. Save raw tool output to files with options such as -oA as described in Note-Taking and Evidence. Then move to deeper scanning in Port Scanning in Depth.

Defender takeaways

  • Do not rely on blocking ping as a security measure; it hides little and complicates troubleshooting.
  • Segment networks so a single compromised host cannot map everything.
  • Monitor for sweeps: many connection attempts to sequential addresses are a classic signal.
  • Keep an asset inventory so unknown live hosts are noticed.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. Why use more than one probe type during host discovery?

    Firewalls may drop ICMP while allowing TCP traffic. Combining probes reduces false negatives.
  2. 2. Which probe is most reliable on the same local subnet?

    ARP is how Ethernet hosts find each other, and it cannot be blocked by a host firewall the way ping can.
  3. 3. What does the nmap option -sn do?

    -sn means no port scan; only discovery.
  4. 4. Every address in a range replies as up. What should you suspect?

    Devices in the path can answer for addresses that have no host, producing false positives.

Summary

  • Host discovery finds which addresses are live before deeper scanning.
  • Probes include ICMP echo, ARP, TCP SYN or ACK to common ports and UDP; combine them.
  • ARP works only on the local subnet; external sweeps depend on firewall behavior.
  • Use polite timing, exclude fragile systems, log commands and times, and stay in scope.
  • Watch for false negatives (silent hosts) and false positives (proxies answering).

Next: Port Scanning in Depth