Feature #6831
Updated by Brandon Murphy 10 months ago
Use Case:
Consider the following HTTP request. I would like to write detection logic that ensures the 32 byte string that occurs in the URI is also found within the Cookie Value.
<pre>
GET /example/path?foo=Pn5tRZrj12eSWFx4qL7cAeyzvKl90O6G HTTP/1.1
Host: example.com
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:97.0) Gecko/20100101 Firefox/97.0
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8
Accept-Language: en-US,en;q=0.5
Accept-Encoding: gzip, deflate
Connection: keep-alive
Referer: http://example.com
Cookie: foo=Pn5tRZrj12eSWFx4qL7cAeyzvKl90O6G
</pre>
reference: https://forum.suricata.io/t/byte-extract-byte-test-string-limits/4511/3
Current and Attempted Methods:
Currently the only way (with PCRE capture groups)to accomplish this detection logic has many limitations/considerations
1) within a single buffer (http.start works in this example)
- this won't work with HTTP/2 traffic
2) via capture groups within PCRE. This can be pretty costly.
I attempted to use the byte_extract/byte_test, which did work, but was limited to 20 byte as noted within the discourse convo. the keyword performance of the byte_extract/byte_test combo was about half as many ticks as using the PCRE capture groups.
Proposed Solution:
I won't bother with suggesting how this should be solved, but more details of the use case are required please let me know.
P.S. - I'm not sure what to "call" this request, so feel free to change the subject to more accurately reflect whatever this is.