> -----Original Message-----
> From: [email protected] [mailto:rsyslog-
> [email protected]] On Behalf Of Champ Clark III [Softwink]
> Sent: Friday, October 08, 2010 6:57 PM
> To: rsyslog-users
> Subject: Re: [rsyslog] liblognorm
> 
> > http://cee.mitre.org/
> 
>       Yep.. I've read through that.  I was looking for something more
> "meaty".  :)
>

I understand, but right now I cannot provide more due to NDA's. I am even
going on a border when I begin to write the class definitions (and so I will
check with Mitre first).
 
> > That sounds good. The only thing that I am pretty sure about is that
> - at
> > some stage - we must support *multiple* files. That is because I
> envision
> > that some may be pulled from a global repository but some local-only
> may also
> > exist. I think it is easier to manage those if they can be kept in
> different
> > files.
> 
>       That's a interesting concept,  and pretty much how we do it with
> Sagan/Snort.  In the Sagan configuration file,  you have lines like
> this:
> 
> include $RULE_PATH/rsync.rules
> include $RULE_PATH/samba.rules
> include $RULE_PATH/sendmail.rules
> 
>       If you don't use "sendmail",  you can "# out" that rule.
> There's not much need to "monitor" for things that you don't expect to
> see.
> The same could apply to liblognorm ... That way,  you could also
> include
> "local" definitions.

Yup

> 
>       Here's how I'm looking to use something like liblognorm.  I'd
> actually already started on some simple parsers,  but would rather see
> something like liblognorm (keeps from re-inventing the wheel,  and
> useful for many projects).
> 
>       Take the following "openssh.rules" line:
> 
> alert tcp $EXTERNAL_NET any -> $HOME_NET 22 (msg:"[OPENSSH]
> Authentication failure for root"; content: "Authentication failure for
> root"; classtype: unsuccessful-admin;program: sshd; threshold:type
> limit, track by_src, count 5, seconds 300; parse_ip_simple;
> parse_port_simple; reference:
> url,wiki.softwink.com/bin/view/Main/5000017; sid: 5000017; rev:1;)
> 
>       Note the parse_ip_simple and parse_port_simple.  Those are my
> current,  simple,  parsers to pull IP address and TCP source port
> information (when applicable).    Replace those calls with the
> liblognorm.   That's the goal,  across many different log sets (Cisco,
> Fortigate firewalls,  Linux boxes) that I'm looking for.
> 
>       Basically, "ip_parse_simple" becomes a rule flag I can pass to
> liblognorm,  which "tells" liblognorm",  "this is a openssh message"
> and
> "extract the source IP address and source port".
> 
>       Does that seem on track?  Sorry for the rant....

Exactly the same idea. Maybe we can share some of the code!

Rainer
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com

Reply via email to