Re: Whitelist certain hits in prelude-lml
| Newsgroups | gmane.comp.security.ids.prelude.user |
|---|---|
| Message-ID | <OFBE61CBD0.B4D3536E-ON0A257540.006E144C-0A257540.006F5DFD@thdfsg.com> |
Date: Mon, 5 Jan 2009 10:56:09 -1000 From: [email protected] Subject: [prelude-user] Whitelist certain hits in prelude-lml To: [email protected] Message-ID: <OFE06ABB60.4C86DCAE-ON0A257535.0072F759-0A257535.007303C7@thdfsg.com> Content-Type: text/plain; charset="US-ASCII" I have a custom rule that I'm trying to write for prelude-lml. I have the regex working, but there is a particular false-positive that I'd like to screen out. How would I set things up so that there wouldn't be a trigger if the target IP is 10.0.0.5? How about if part of the regex matches a certain string? (IE, skip if $4 = "False Alarm") To answer my own question: In the relevant ruleset, place an "emtpy" match above the one being triggered with a more specific regex. For example, suppose you have a firewall type rule that matches "Alarm" and captures the source ip, destination ip and desitination port, but you want to skip it if the source IP is 10.10.10.10 and destination port is 1001 #### Trigger on "Alarm: " #### Unless it's on port 1001 from source IP 10.10.10.10 regex=Alarm: .*, src: 10.10.10.10, ([\d\.]+), proto: tcp, port: 1001; \ last regex=Alarm: (.*), src: ([\d\.]+), ([\d\.]+), proto: tcp, port: (\d+); \ id = 9999; \ revision= 9; \ analyzer(0).name = foobar analyzer(0).class = something assessment.impact.type = whatever ... rest.of.the.stuff.goes = here ... last Dean Takemori Tech Support Supervisor TD Food Group [email protected] _______________________________________________ Prelude-user site list [email protected] http://lists.prelude-ids.org/mailman/listinfo/prelude-user