Lion Cipher Lab

Triage Snort IDS alerts: signatures, source IPs, and separating attacks from noise

Why it matters

A sensor that fires is only half a detection capability; the other half is a human who can look at the output and say, in under a minute, 'this is an attack on that host, this is a scanner, and this one is just a Linux box patching itself.' That reading skill is the day-in-day-out work of a SOC analyst, and it is what an interviewer is really testing when they slide a screen of alerts across the table and say 'walk me through this.' The trap is treating every line as its own event: real incidents live in the relationship BETWEEN lines — the same source IP that scanned you at 09:14 is the one…

What you'll learn

Steps in this lab

  1. What this drill is, and the one idea to hold onto
  2. The alerts you are triaging — read the whole block once
  3. Anatomy of one alert line, left to right
  4. The gid:sid:rev — and how to look a signature up
  5. classtype and priority — what the sensor thinks it means, and how much to trust that
  6. Group the burst by source IP — five lines, one story
  7. Read the sequence — recon, then exploit, then the callback that proves it
  8. Separate the true positives from the noise — spot the false positive
  9. When a signature is enough to act — and when you must pull the packet
  10. Write the verdict, and how to do this better than the naive way

Open the full interactive lab →