Your NGFW Has 10,000 Signatures — How Many Actually Fire?

IPS enabled. Signatures updated. Dashboard green. But when real attack traffic crosses the wire, how many of those signatures actually trigger?

March 15, 2026·9 min read·NGFW

The Configuration Gaps Nobody Tests For

We see the same pattern everywhere: a six-figure NGFW investment, thousands of IPS signatures licensed, and a dashboard that shows everything is healthy. Then someone runs an actual attack test and half the signatures don't fire.

SSL decryption is disabled — the firewall can't see inside HTTPS traffic, making every IPS signature useless against 95% of web traffic. IPS is in detect-only mode — someone switched it during troubleshooting and never switched it back. Signature databases are stale — the license expired or auto-update was never configured. Rule exceptions for "trusted" apps — permanent holes in your inspection policy for applications decommissioned quarters ago.

Combine two or three of these, and your NGFW is an expensive packet forwarder. The only way to know is to send real attack traffic and see what happens.

Browser-Based Testing Over HTTPS — Why That Matters

Most NGFW testing tools require agents, appliances, or paid subscriptions — BreakingPoint, Keysight, Spirent. They're powerful but cost more than many organizations' entire security budget.

ITSecTools runs entirely from your browser. Navigate to the URL, click a button, and real attack payloads flow through your network as HTTPS traffic. This is critical because testing over HTTPS simultaneously validates that SSL decryption is working. If your test tool only sends attacks over HTTP, you get a false sense of security — your IPS blocks HTTP attacks fine, but encrypted attacks sail through because nobody verified the decryption policy.

Prerequisite: Your firewall must have SSL decryption (DPI-SSL) enabled for itsectools.com. Without it, the firewall sees encrypted blobs and IPS signatures never trigger. If everything shows as "allowed," check your decryption policy first.

4 Test Suites That Cover the Full Attack Surface

IPS Signature Testing (3 tests) — SQL injection, XSS, and directory traversal in HTTP query strings. The bread-and-butter signatures every IPS should catch. If these pass, you have a fundamental problem.

Advanced Evasion Techniques (3 tests) — Log4j JNDI in HTTP headers (not the URL), hex-encoded SQL injection (tests content normalization), and Shellshock RCE in custom headers. Catches firewalls that only do single-pass pattern matching.

C2 Beacon Simulation (3 tests) — Outbound data exfiltration, web shell beacon check-ins, and ActiveX dropper delivery via response body. Blocking the initial exploit is half the job — if the C2 callback gets through, the attacker still wins.

Network IP Flooder (30 attacks) — 13 unique attack patterns (SQL injection variants, 5 path traversal encoding methods, system file disclosure), randomized with Fisher-Yates shuffle, fired rapidly with no delay. Tests whether your firewall maintains accuracy under sustained volume.

[Screenshot: NGFW Validation interface showing all 4 test suites with IP shun cooldown configuration]

Mid-Stream Body Termination Detection

Here's something unique to ITSecTools. When an NGFW performs streaming content inspection, it sometimes allows the HTTP 200 headers through but kills the connection mid-stream when it detects malicious content. Most testing tools count HTTP 200 as "allowed." ITSecTools reads the full response body and catches mid-stream termination as a block — because that's exactly what it is.

This detection method was critical in achieving accurate flood test results. Without it, many legitimate blocks get misreported as successful attacks.

What Your Results Tell You

IPS blocks but AET passes: Single-pass pattern matching without content normalization. Enable decoding/normalization in your IPS policy.

IPS + AET block but C2 passes: Inbound threat prevention works, but outbound egress filtering is weak. An attacker who gets in can phone home freely.

Everything passes: SSL decryption is almost certainly disabled. Enable DPI-SSL and retest.

Flooder: 30/30 blocked but only 3-5 NGFW logs: Normal. After first detection, your firewall shuns the source IP — subsequent attacks drop at kernel level without individual logs. All 30 are blocked; logs reflect shunning behavior.

[Screenshot: Console output showing color-coded results across IPS, AET, C2, and Flood suites with blocked/allowed summary]
Validate Your NGFW →