derez on 22 Oct 2005 14:51:03 -0000 |
Brian Hawkins wrote: > >I'm intending on using tcptraceroute to do some mapping of the >internet for a class I'm taking and I don't want my activity to be seen as an >attack on a server. There are a few known SYN attacks were >the attacker tries to fill up the servers connections, but if what you say is true >and the computer running tcptraceroute sends back an RST this >should not be a problem. Might I offer that you take the tool and run it in a test environment with a packet sniffer to see what it does. It is always helpful to have a full understanding of what it is doing especially when used for research. In addition run a packet sniffer when using it as you may see wierd situations and having additional data could help with analysis. I have seen strange behaviour when dealing with layer 7 switches, load balancers, homemade firewalls, etc. and having the additional data was helpful. As far as the number of connections initiated, the operator has control over that with the number of queries (-q) argument. I usually stick with the default (3) as sometimes a particular hop takes time to find which can be somewhat verified with the first response time that is larger than the other two reponse times. ~derez _______________________________________________ tcptraceroute-dev mailing list tcptraceroute-dev@netisland.net http://lists.netisland.net/mailman/listinfo/tcptraceroute-dev
|
|