|
Nmap Development
mailing list archives
Re: Desired improvements in Nmap performance? [SCAN BUDDIES]
From: Brandon Enright <bmenrigh () ucsd edu>
Date: Wed, 3 Dec 2008 00:14:21 +0000
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Tue, 2 Dec 2008 17:00:08 -0700
David Fifield <david () bamsoftware com> wrote:
Now with the --initial-rtt-timeout:
$ sudo ./nmap --datadir ./ -p- -T5 -v -d -PN -n
--initial-rtt-timeout 50 132.239.7.132 SYN Stealth Scan Timing:
About 0.18% done Current sending rates: 33.81 packets / s, 1472.89
bytes / s.
Of course, the buddy was *much* faster than this.
I see, it also has to do with the congestion window (number of probes
allowed to be outstanding at once). Just like with the RTT, hosts with
no responses can take their congestion window from the global group
(HostScanStats::getTiming). The scan buddy allows the global
congestion window to grow. My guess is that combining
--min-parallelism with --initial-rtt-timeout would give you
comparable speeds. With -d3 you can get an output of the current
congestion window size. It looks like
**TIMING STATS** (0.7920s): IP, probes
active/freshportsleft/retry_stack/outstanding/retranwait/onbench,
cwnd/ccthresh/delay, timeout/srtt/rttvar/ Groupstats (1/1
incomplete): 49/*/*/*/*/* 49.75/75/* 154278/66678/21900
Here the congestion window has a size of 49.75.
David Fifield
Bingo. I ran:
$ time sudo nmap -v -d3 -p- -T5 -PN -n 132.239.7.132,131 --open 2>&1 | egrep -A 1 'TIMING STATS' | tail -n 2
The last stats were:
**TIMING STATS** (25.0050s): IP, probes active/freshportsleft/retry_stack/outstanding/retranwait/onbench,
cwnd/ccthresh/delay, timeout/srtt/rttvar/
Groupstats (1/2 incomplete): 0/*/*/*/*/* 300.00/237/* 50000/268/26
I then ran (without the buddy):
$ time sudo nmap -v -d3 -p- -T5 -PN -n 132.239.7.132 --initial-rtt-timeout 50 --min-parallelism 300 --open 2>&1 | egrep
-A 1 'TIMING STATS' | tail -n 2
The scans finished within a second of each other.
Brandon
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)
iEYEARECAAYFAkk1z2MACgkQqaGPzAsl94I9BgCcDNHy8FIqNQ0KNcY5SmWrdqO3
hfYAn2GAN8xj86CbiKsYqgPSdjqEBApo
=QjJN
-----END PGP SIGNATURE-----
_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://SecLists.Org
By Date
By Thread
Current thread:
- Re: Desired improvements in Nmap performance?, (continued)
Re: Desired improvements in Nmap performance? [FASTER IS SLOWER] Brandon Enright (Dec 02)
Nmap performance work update David Fifield (Dec 08)
Re: Desired improvements in Nmap performance? Rob Nicholls (Dec 01)
|