Status Report #15 of 17

[Joao Correa <joao () livewire com br>]

Accomplishments:

- Finished some modularization improvements in HTTP

- Many improvements in HTTP Pipelining:
-- Transparently deal with servers that do not support pipelining
-- Number of requests change dynamically depending on results received
-- Fixed maximum number of simultaneous pipeline requests
-- Implemented HEAD pipeline

- Implemented sql-injection using pipeline and made comparison between
both implementations*

- Fixed a bug in sql-injection that leaded the script to discard
positive results (due to 500 status code, internal server error not
being considered a valid result)

- Made possible to pass the maxdepth as argument to sql-injection

- Applied pipelining to http-enum.nse

Priorities:

- Write a "pipeline aware" request function, that is able to handle
with timeout and make better use of the open socket.
- Check for new priorities on tomorrow's meeting

* Indeed pipelining provided very good results. I've tested it with a
small set of web applications so far, and, for the sql-injection.nse
script, I could notice that the performance was not harmed by the
pipeline when pipelining was not supported. With pipeline support I've
noticed an amazing performance increase: A scan that usually takes
11804 secs (more than 1100 malicious requests) was done with only 2090
secs.

_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://SecLists.Org