Reading Logs, Verifying Hashes, and Detecting Data Attacks

4.4.ASkill

Reading Windows Security Event Logs

Windows logs authentication activity with standardized numeric Event IDs. Two are foundational enough to know on sight:

📋 Event ID Reader — click each entry
SUCCESS
A successful logon. On its own this is routine — but a 4624 from an unusual location, device, or time is worth flagging.
🔑A single 4625 means nothing on its own. Ten 4625 entries for the same account within one minute, followed by a 4624 — that's the signature of a successful brute-force or credential-stuffing attempt, and it's exactly the pattern you're trained to notice now.
5.6.A–CConcept

Detecting Attacks on Data

Devices track and log when data is accessed and by whom — a process called accounting. Reviewing those logs can surface malicious activity: files accessed outside someone's normal pattern, unusual login times or locations, or attempts to copy or delete sensitive files.

ToolHow it worksCost
HoneypotA fake file that looks valuable — any access at all is suspicious, since there's no legitimate reason to touch itLow
DLP ServiceMonitors data access, usage, and transmission org-wideHigher
Hash VerificationDetects whether a file's contents changedLow
⚠️Every one of these has a blind spot. A honeypot can't catch an adversary who never touches it. A hash can't catch data that was viewed or copied without being changed. Real detection strategies usually combine multiple tools precisely because no single one covers every case.
5.6.DSkill

Verifying Integrity with File Hashes

Cryptographic hash functions are repeatable — the same file always produces the same hash. That property is what makes hash verification work: hash a file, record the result, and hash it again later. If the two hashes match, the file is provably unchanged.

🔐 Hash Verification Workflow
Get-FileHash evidence.docx -Algorithm SHA256
Output: 8f3a...c92d
💡MD5 vs. SHA-256: MD5 is faster but cryptographically weaker — collisions (two different files producing the same hash) have been demonstrated. SHA-256 is slower but far more collision-resistant, which is why it's the standard choice when hash verification needs to hold up as forensic evidence.
4.4.CConcept

Evaluating a Detection Method

Choosing a detection method isn't just "pick the best one" — it's a real tradeoff across three factors:

FactorThe tradeoff
Speed / performanceSignature-based is faster; anomaly-based can degrade performance, especially on limited-power devices
Attack phaseCatching an attack earlier (at the device level) prevents more damage than catching it after data is already gone
False positives vs. bypass easeSignature-based has fewer false alarms, but is easier for adversaries to evade than behavior-based detection
5.6.ESkill

Spotting SQL Injection in a Log

SQL injection attempts leave fingerprints in application and server logs — specific characters and patterns that shouldn't appear in normal input.

ExampleGuided Example — Scanning a Log for Injection Attempts

A web server log shows this input submitted to a login form's username field: username=admin' OR 1=1 --

Step 1Spot the quote character
The single quote (') is trying to close the string the application expected — the first sign this isn't a real username.
← Back to Activity 2.2.3Next: Project 2.2.4 →Secure the Server.