Skip to content
HACKTECH

Library / Practice Lab

Build a range you can safely break.

A segmented lab for red, blue and purple work: the machines, the network boundaries, the scenarios to run, and — first — the rules that keep it from ever touching a network you did not mean to.

Read this before you build

  • The lab must not route to your home, office or any production network. Verify this before you install a single tool.
  • Use host-only or internal virtual networks. Give the attack workstation internet access only when installing tooling, then remove it.
  • Never point lab tooling at an address you do not own. A typo in a scan range is how people end up in serious trouble.
  • Handle malware samples only on an isolated segment with no egress, on a snapshot you intend to discard.
  • Use evaluation licences for Windows images and rebuild when they expire rather than tampering with activation.
  • Snapshot every node in a known-good state before each exercise. You will need it more than you think.

The machines

Attacker

Attack workstation

Kali Linux or Parrot Security

4 vCPU · 8 GB RAM · 80 GB disk

Your operating position. Everything offensive originates here so that traffic attribution stays unambiguous.

Snapshot immediately after tooling is installed and configured. You will want to roll back to a clean state more often than you expect.

Infrastructure

Domain controller

Windows Server 2022 (evaluation)

2 vCPU · 4 GB RAM · 60 GB disk

Active Directory, DNS and DHCP for the lab domain. The centrepiece of any realistic internal scenario.

Deliberately introduce the misconfigurations you want to study: Kerberoastable service accounts, weak ACLs, unconstrained delegation.

Target

Domain workstation

Windows 11 Enterprise (evaluation)

2 vCPU · 4 GB RAM · 60 GB disk

A domain-joined endpoint representing initial foothold. Where credential access and lateral movement begin.

Install Sysmon with a well-tuned configuration so you can see exactly what your own techniques generate.

Target

Linux application server

Ubuntu Server LTS

2 vCPU · 4 GB RAM · 40 GB disk

Hosts deliberately vulnerable web applications for the web track: Juice Shop, DVWA, WebGoat.

Run each application in its own container so you can reset one without disturbing the others.

Monitoring

Monitoring & SIEM

Ubuntu Server LTS

4 vCPU · 8 GB RAM · 120 GB disk

Collects logs from every node. This is what turns an attack lab into a detection lab.

Wazuh or an Elastic stack both work well. The disk fills faster than you plan for — size generously.

Infrastructure

pfSense firewall

pfSense CE

2 vCPU · 2 GB RAM · 20 GB disk

Segments the lab and controls egress. Also gives you a realistic target for segmentation testing.

Configure it so lab traffic cannot reach your home or office network. This is the single most important step on the page.

Network segments

SegmentCIDRPurposeRules
Management10.10.0.0/24Hypervisor and firewall administration only.Reachable from your host. No inbound from any other lab segment.
Attack10.10.10.0/24Attack workstation and any C2 infrastructure.Full outbound to Target and DMZ. No route to Management.
Corporate10.10.20.0/24Domain controller and domain workstations.Simulates an internal network. Logs everything to Monitoring.
DMZ10.10.30.0/24Internet-facing application servers.Reachable from Attack. Constrained path into Corporate, so you can practise pivoting.
Monitoring10.10.40.0/24SIEM and log collection.Accepts inbound log traffic from all segments. Initiates nothing outbound.

Scenarios to run

External to domain admin

Advanced · 4–8 hours

Start from the DMZ web application with no credentials. Reach domain administrator on the Corporate segment.

  • Exploit the DMZ application to obtain a foothold
  • Discover credentials reusable against the Corporate segment
  • Enumerate the domain and identify an escalation path with BloodHound
  • Escalate to domain administrator and document every step
  • Review the SIEM afterwards: which steps generated telemetry, and which did not?

Detection engineering loop

Intermediate · 3–5 hours

Run techniques from Atomic Red Team, then write the detection that catches each one.

  • Execute a chosen ATT&CK technique from the attack workstation
  • Identify the telemetry it produced in Sysmon and the SIEM
  • Write a Sigma rule that detects it
  • Re-run and confirm the rule fires
  • Deliberately vary the technique and see whether your rule survives

Segmentation validation

Intermediate · 2–3 hours

Verify the firewall rules do what you believe they do. This is a real client deliverable, not just an exercise.

  • Document the intended policy between every pair of segments
  • Test each path systematically from both directions
  • Find at least one rule that behaves differently from its description
  • Produce a matrix of intended against actual reachability

Incident response tabletop

Foundation · 2 hours

Have someone else compromise the lab, then work the incident from the defender's side using only the SIEM.

  • Establish the earliest evidence of compromise from logs alone
  • Reconstruct the attacker's path without asking them
  • Identify what evidence you wish you had collected and were not
  • Write the first-hour timeline as if it were a real incident record

Vulnerable apps to host

  • OWASP Juice Shop

    Modern JavaScript application with a very wide challenge set. The best all-round target.

  • DVWA

    Classic PHP application with adjustable difficulty. Good for teaching one class at a time.

  • WebGoat

    OWASP's guided lessons with explanation built into each exercise.

  • VulnHub

    Downloadable vulnerable machine images spanning every difficulty level.

  • GOAD

    Game of Active Directory: a prebuilt multi-domain lab with intentional misconfigurations. Saves days of setup.

  • DetectionLab

    Automated build of a Windows domain with logging and monitoring preconfigured.