Skip to content
Notifications
Clear all

Has anyone quantified the latency added by AWS WAF on an ALB?

46 Posts
43 Users
0 Reactions
506 Views
(@ethan9)
Estimable Member
Joined: 4 months ago
Posts: 194
 

I've conducted several structured benchmarks for e-commerce platforms, and the 2-4ms baseline for the Core Rule Set is reproducible under controlled conditions. However, your question about variation between rule sets is crucial. The delta between the Core Rule Set and a complex custom configuration isn't linear; it's multiplicative on specific request types.

The heaviest impact you'll see is from custom regex rules that evaluate the request body. For a checkout flow POST with a JSON payload, I've measured a single, complex regex rule adding 14ms at p95, while the entire managed rule set added only 3ms. The distribution is the key - the average increase might be 5ms, but your checkout p99 could jump 20ms.

Since you're from a data quality background, structure your test like a controlled experiment. Isolate the WAF latency by enabling full logging and extracting the `waf_evaluation_time` field from the ALB logs. Script your exact checkout sequence in k6, and add rules in phases: 1) Core Rule Set only, 2) Core + rate limiting, 3) Core + your proposed custom rules. Measure the delta for each phase specifically on the POST endpoint. This will give you the marginal cost of each security layer for your critical path.


Data never lies.


   
ReplyQuote
Page 4 / 4