Module 19 Lesson 6 of 6 🕑 ~80 min

> cat module-19-6-documentation-automation-capstone.md

Documentation, Automation & the Capstone

The environment is built. Now it needs to survive contact with failure, get documented well enough that someone else could understand it, and turn into something you can actually put in front of an interviewer — which is the whole point of this module.

1 Breaking the Lab & Example Incidents

Deliberately create faults: stop DNS, use the wrong DNS server, disable DHCP, block Kerberos, expire a certificate, lock an account, disable a user, stop SQL Server, stop the web server, block TCP 443, fill a disk, change file permissions, stop the Splunk forwarder, remove a route — then troubleshoot each one systematically. This may be the single most educational part of the entire lab.

Users report https://portal.corp.cybersecuritycafe.lab is unavailable. What's your step-by-step investigation?

Reveal the layered approach

1. Client: can it resolve the hostname (nslookup portal.corp.cybersecuritycafe.lab)? 2. Network: can it reach the server — ping where permitted, check TCP connectivity. 3. Firewall: is TCP 443 allowed? 4. Server: is the web server running? 5. Certificate: is TLS valid? 6. Application: is the application actually running? 7. Database: can the application reach the database? 8. Logs: what changed? This is layered troubleshooting exactly as Module 16 teaches it — work outward from the client, one layer at a time, instead of guessing.

A user cannot log in. What's the full range of possible causes, before you assume it's a password problem?

Reveal the checklist

Username, password, account disabled, account locked, password expired, workstation network, DNS, Domain Controller availability, Kerberos, time synchronisation, and Group Policy — and possibly more. "Login problem" can describe many completely different failures, which is exactly why the six-step authentication troubleshooting flow from Module 16 exists.

2 A Security Investigation & Vulnerability Management

Generate abnormal behaviour — many failed logins followed by a successful one — and send the logs to Splunk or Sentinel. Investigate the user, timestamp, source device, source address, number of attempts, whether authentication succeeded afterward, and any other activity involving the same account. Document the investigation as if you were actually working in a SOC (Module 11's investigation format applies directly here).

Use appropriate vulnerability-scanning tools against systems inside your own controlled lab and identify missing patches, weak configurations, unnecessary services, open ports, outdated packages, and TLS configuration problems. The important part is what happens after the scan: for every finding, ask what's vulnerable, why it matters, whether it's exploitable in this specific architecture, what the remediation is, whether remediation could break anything, and how the fix will be tested. That's what separates real vulnerability management from just producing a scan report nobody acts on.

3 Documentation, Diagrams & the GitHub Portfolio

Document everything: architecture, IP addresses, hostnames, operating systems, applications, firewall rules, user roles, service accounts, dependencies, backup process, and troubleshooting notes. A simple inventory table goes a long way:

HostnameFunctionOSIP
FW01FirewallpfSense10.10.10.1
DC01Domain ControllerWindows Server10.10.10.10
DC02Domain ControllerWindows Server10.10.10.11
FILE01File ServerWindows Server10.10.10.20
SQL01DatabaseWindows/Linux10.10.10.30
LINUX01Linux ServerUbuntu/Rocky10.10.10.40
SPLUNK01SIEMLinux10.10.10.50
WIN11-01WorkstationWindows 11DHCP
KALI01Security workstationKali Linux10.10.30.10

Create a professional network diagram (draw.io, Visio, Lucidchart, Mermaid or Excalidraw all work) — being able to explain architecture visually is a real professional skill, not decoration.

Then publish a repository, e.g. enterprise-home-lab, structured with a README.md, a /docs folder (architecture, active-directory, networking, databases, logging, security), a /diagrams folder, a /scripts folder, and a /screenshots folder. Never upload passwords, API keys, private keys, cloud credentials, or real company information — everything in this lab is fictional by design, keep it that way.

Write the README as an actual pitch, not a file listing:

Enterprise IT & Cybersecurity Home Lab

Built a segmented enterprise-style network containing:
- Windows Server, Active Directory, DNS/DHCP
- Windows clients, Linux, pfSense
- SQL Server, Docker, Splunk, Microsoft Sentinel
- Kali Linux, vulnerable applications

Implemented:
- RBAC, Group Policy, network segmentation
- Database access controls, centralised logging
- SIEM monitoring, certificate-based HTTPS
- Database security, backup and recovery

This is far more valuable than "I know Active Directory" on a CV, because every line is something a reviewer can ask you to explain in detail.

4 Automation: PowerShell & Linux Scripting

Automate part of the build. Use PowerShell to create users from a CSV of dozens of fictional employees:

Import-Csv users.csv | ForEach-Object {
    New-ADUser `
        -Name $_.Name `
        -SamAccountName $_.Username `
        -Enabled $true
}

Now you're demonstrating Active Directory, PowerShell, CSV handling and automation in a single project (Module 15 covers PowerShell in depth). On the Linux side, write a basic shell script to collect hostname, disk usage, memory, IP address, and running services, and schedule it with cron — connecting the Linux and automation skills from earlier in the course.

5 Enterprise Scale & the Recommended Final Lab

Your lab might have 10 servers, 5 users, and 2 networks. A real enterprise might have 20,000 servers, 80,000 employees, 200 offices, multiple data centres, AWS, Azure, GCP, thousands of applications, multiple Active Directory forests, and multiple SIEM platforms. The underlying concepts stay the same — what changes at scale is the need for automation, standardisation, monitoring, governance, high availability, change management, security controls, and documentation. A realistic organisation might also simultaneously run Windows Server 2016 through 2025, Active Directory alongside Microsoft Entra ID, NTLM alongside Kerberos, OAuth 2.0, OIDC and SAML, Oracle and SQL Server and PostgreSQL, VMware, Docker and Kubernetes, and on-premises infrastructure alongside cloud and SaaS — and that complexity is completely normal, not a sign something's wrong.

By the end of this module, aim for: FW01 (pfSense), DC01 and DC02 (Windows Server/AD/DNS), FILE01 (Windows file server), WIN11-01 and WIN11-02 (Windows 11), LINUX01 (Linux server), DOCKER01 (Linux + Docker), SQL01 (SQL Server/PostgreSQL/MySQL), optionally ORACLE01, SPLUNK01 (Splunk Enterprise), KALI01 (Kali Linux), VULN01 (vulnerable applications) — plus an Azure presence with Microsoft Sentinel and Microsoft Entra ID, and Server/Client/Security VLANs plus a DMZ.

6 The Capstone Project

Scenario: Cyber Security Cafe Ltd has hired you as a junior infrastructure and security engineer. Build an Active Directory domain with two Domain Controllers, Windows workstations, a Linux server, a corporate file server, a SQL database, a Docker application, a firewall, network segmentation, centralised logging, a SIEM, and a security testing environment. Create departments (IT, Finance, HR, Sales, Security) and at least 20 fictional employees, then implement RBAC, Group Policy, file permissions, database permissions, firewall segmentation, administrative accounts, service accounts, logging, monitoring, and backups. What order would you build this in, and why?

Reveal a sensible build order

Roughly the order this module taught it in, because each layer depends on the one before: firewall and networks first (nothing else has anywhere to live otherwise), then Active Directory and DNS/DHCP (everything else needs identity and name resolution), then file services, Linux, database and Docker, then logging/SIEM once there's something worth monitoring, then the DMZ and public-facing pieces last, hardened from day one rather than bolted on afterward. Deviating from this order is fine once you understand why it's the default — the point is having a reason, not memorising a sequence.

Capstone incident: a user reports they can't access the corporate application. Investigation eventually reveals the application server can't communicate with the database server — a firewall change accidentally blocked the database port. Walk the 12-step response.

Reveal the response sequence

Confirm the incident, determine scope, check DNS, check connectivity, check the application service, check the database service, review firewall logs, identify the blocked connection, correct the firewall policy, test the application, document the incident, and document preventative actions. Completing this demonstrates networking, Windows, Linux, database knowledge, firewall troubleshooting, application troubleshooting, logging, incident handling, communication, and documentation — effectively the entire course, compressed into one scenario.

Capstone security incident: you generate several authentication failures against a test account and receive a SIEM alert for multiple failed authentication attempts. What does the resulting incident report need to contain?

Reveal the report structure

Investigate the account, source machine, source IP, timestamp, failure reason, whether authentication succeeded afterward, and any other activity involving the account. Then write it up with an executive summary, timeline, evidence, investigation, impact, containment, root cause, and recommendations (Module 12's incident report structure applies directly). This is something you can genuinely discuss in a SOC interview, because you actually generated the alert, investigated it, and wrote the report yourself.

7 Interview Prep & the Final Challenge

After completing the lab, you should comfortably answer questions like: what happens when a computer joins an Active Directory domain? What's a Domain Controller, and why would an organisation have two? Why is DNS important to Active Directory? What's the difference between NTLM and Kerberos? What's RBAC, and what's a service account? How would you troubleshoot a user who can't access a file share? What's network segmentation, a DMZ, a VLAN? Why shouldn't databases normally be directly accessible from every workstation? What's a SIEM, an index, KQL? What's the difference between a VM and a container? How do SSH keys work? Why shouldn't an application connect as a database administrator? How would you troubleshoot HTTP 500 vs. 401 vs. 403, or an application that suddenly becomes unavailable? A student who genuinely built and broke this environment answers these with practical examples, not memorised definitions.

Rewrite your CV accordingly. Instead of "Knowledge of Windows Server and cybersecurity," write something closer to: "Designed and deployed a segmented enterprise-style infrastructure using Windows Server, Active Directory, DNS, DHCP, Linux, pfSense, SQL Server, Docker, Splunk and Microsoft Sentinel. Implemented Active Directory RBAC, Group Policy, network segmentation, database access controls, centralised logging and security monitoring. Created Splunk and Sentinel detections for authentication activity and performed simulated incident investigations. Automated Active Directory administration using PowerShell and documented infrastructure using network diagrams, configuration documentation and troubleshooting runbooks." Every sentence is now something an interviewer can ask you to go deeper on. When they ask "Do you have any experience with Active Directory?", instead of "I studied Active Directory," you can say exactly what you built, what you configured, and what you tested when things failed — that difference is enormous.

Final Challenge — No Step-by-Step

Build the environment from this requirement alone, with no tutorial to follow: "Cyber Security Cafe Ltd requires an internal IT environment supporting 50 employees across Finance, HR, IT, Security and Sales. Employees require Active Directory authentication, Windows workstations, shared storage, an internal web application, a database, centralised logging and security monitoring. Internet-facing services must be separated from internal infrastructure. Administrators must use separate privileged accounts. Security events must be searchable centrally." Decide for yourself which servers and networks are required, where the database and Domain Controllers should live, what firewall rules and security groups are needed, how users should authenticate, which logs to collect, what to monitor, how to back up, how administrators should connect, and what happens if a server fails.

This is where you move from following labs to designing systems — one of the biggest steps toward becoming employable in IT and cybersecurity.

Module 19 Complete — and Section 04

You should now be able to open a laptop and show a real network: Domain Controllers, users, Group Policy, a Linux server, a database, a Docker environment, a firewall and its rules, a Splunk dashboard, Sentinel queries, an incident you investigated, PowerShell automation you wrote, a network diagram, documentation, something you broke, how you diagnosed it, and how you fixed it. That's practical experience — and it's worth more to an employer than a certificate saying you completed a course.

The lab doesn't need to be huge. It needs to demonstrate that you understand how modern and legacy enterprise technologies fit together, know how to troubleshoot and secure them, and can explain what you built and why. Mark this lesson complete to close out Module 19 and Section 04: Professional Communication, and continue to Section 05: Getting Your First Job.