Commit d0e0af4d authored by Devon Kearns's avatar Devon Kearns

Imported Upstream version 0.30

Main author:
pagvac (
Patches, bug hunting and general feedback:
Borys Lacki (
Philipp Winter (
meathive (
David Kierznowski (
GNa (
srl (
20/02/2010: dnsmap 0.30 released at
* IPv6 support
* Makefile included
* delay option (-d) added. This is useful in cases where dnsmap is killing your bandwidth
* ignore IPs option (-i) added. This allows ignoring user-supplied IPs from the results. Useful for domains which cause dnsmap to produce false positives
* changes made to make dnsmap compatible with OpenDNS
* disclosure of internal IP addresses (RFC 1918) are reported
* updated built-in wordlist
* included a standalone three-letter acronym (TLA) subdomains wordlist
* domains susceptible to "same site" scripting ( are reported
* completion time is now displayed to the user
* mechanism to attempt to bruteforce wildcard-enabled domains
* unique filename containing timestamp is now created when no specific output filename is supplied by user
* various minor bugs fixed
22/02/2009: dnsmap 0.22 released at
* saving the results in human-readable and CSV format for easy processing
* fixed bug that disallowed reading wordlists with DOS CRLF format
* improved built-in subdomains wordlist
* new bash script ( included which allows running dnsmap against a list of domains from a user-supplied file. i.e.: bruteforcing several domains in a bulk fashion
* bypassing of signature-based dnsmap detection by generating a proper pseudo-random subdomain when checking for wildcards
17/08/2006: dnsmap 0.1 (first public version) released at
* bruteforcing based on builtin list and user-supplied wordlist
* obtain all available A records for each bruteforced (sub)domain (rather than only one)
dnsmap: dnsmap.c dnsmap.h
$(CC) $(CFLAGS) -o dnsmap dnsmap.c
install: dnsmap
mkdir -p $(DESTDIR)$(BINDIR)
install -m 0755 dnsmap $(DESTDIR)$(BINDIR)
install -m 0755 $(DESTDIR)$(BINDIR)/dnsmap-bulk
dnsmap was originally released back in 2006 and was inspired by the
fictional story "The Thief No One Saw" by Paul Craig, which can be found
in the book "Stealing the Network - How to 0wn the Box"
dnsmap is mainly meant to be used by pentesters during the information
gathering/enumeration phase of infrastructure security assessments. During the
enumeration stage, the security consultant would typically discover the target
company's IP netblocks, domain names, phone numbers, etc ...
Subdomain brute-forcing is another technique that should be used in the
enumeration stage, as it's especially useful when other domain enumeration
techniques such as zone transfers don't work (I rarely see zone transfers
being *publicly* allowed these days by the way).
If you are interested in researching stealth computer intrusion techniques,
I suggest reading this excellent (and fun) chapter which you can find for
*free* on the web:
I'm happy to say that dnsmap was included in Backtrack 2, 3 and 4 and has
been reviewed by the community:
Compiling should be straightforward:
$ make
$ gcc -Wall dnsmap.c -o dnsmap
Example of manual installation:
# cp ./dnsmap /usr/local/bin/dnsmap
If you wish to bruteforce several target domains in bulk fashion, you can use the
included script. Just copy the script to /usr/local/bin/ so you can
call it from any location. e.g.:
# cp ./ /usr/local/bin/
And set execute permissions. e.g.:
# chmod ugo+x /usr/local/bin/
Lack of multi-threading. This speed issue will hopefully be resolved in future versions.
1. Finding interesting remote access servers (e.g.:
2. Finding badly configured and/or unpatched servers (e.g.:
3. Finding new domain names which will allow you to map non-obvious/hard-to-find netblocks
of your target organization (registry lookups - aka whois is your friend)
4. Sometimes you find that some bruteforced subdomains resolve to internal IP addresses
(RFC 1918). This is great as sometimes they are real up-to-date "A" records which means
that it *is* possible to enumerate internal servers of a target organization from the
Internet by only using standard DNS resolving (as oppossed to zone transfers for instance).
5. Discover embedded devices configured using Dynamic DNS services (e.g.:
This method is an alternative to finding devices via Google hacking techniques
Bruteforcing can be done either with dnsmap's built-in wordlist or a user-supplied wordlist.
Results can be saved in CSV and human-readable format for further processing. dnsmap does
NOT require root privileges to be run, and should NOT be run with such privileges for
security reasons.
The usage syntax can be obtained by simply running dnsmap without any parameters:
$ ./dnsmap
dnsmap 0.30 - DNS Network Mapper by pagvac (
usage: dnsmap <target-domain> [options]
-w <wordlist-file>
-r <regular-results-file>
-c <csv-results-file>
-d <delay-millisecs>
-i <ips-to-ignore> (useful if you're obtaining false positives)
Note: delay value is a maximum random value. e.g.: if you enter 1000, each DNS request
will be delayed a *maximum* of 1 second. By default, dnsmap uses a value of 10 milliseconds
of maximum delay between DNS lookups
Subdomain bruteforcing using dnsmap's built-in word-list:
$ ./dnsmap
Subdomain bruteforcing using a user-supplied wordlist:
$ ./dnsmap -w wordlist.txt
Subdomain bruteforcing using the built-in wordlist and saving the results to /tmp/ :
$ ./dnsmap -r /tmp/
Since no filename was provided in the previous example, but rather only a path, dnsmap would
create an unique filename which includes the current timestamp. e.g.:
Example of subdomain bruteforcing using the built-in wordlist, saving the results to /tmp/,
and waiting a random maximum of 3 milliseconds between each request:
$ ./dnsmap -r /tmp/ -d 300
It is recommended to use the -d (delay in milliseconds) option in cases where dnsmap is
interfering with your online experience. i.e.: killing your bandwidth
Subdomain bruteforcing with 0.8 seconds delay, saving results in regular and CSV format,
filtering 2 user-provided IP and using a user-supplied wordlist:
$ ./dnsmap -d 800 -r /tmp/ -c /tmp/ -i, -w ./wordlist_TLAs.txt
For bruteforcing a list of target domains in a bulk fashion use the bash script provided. e.g.:
$ ./ domains.txt /tmp/results/
OTHER SIMILAR TOOLS - choice is freedom!
Fierce Domain Scan
pagvac |
Feb 2010
* multi-threading - use pthread.h?
* can't handle wildcarded domains that return more than one IP address on non-existing subdomains
test domain:
* allow using a customized list of DNS server to share network load
* allow using DNS server supplied on the command line
* for openDNS users: document how to permanently change DNS server settings so they are not overwritten by DHCP settings
* convert hostent structs to addrinfo ?
* replace inet_ntoa(*((struct in_addr *)host->h_addr_list[j])) with ipstr
* obtain aliases for each domain (CNAME records)?
* clever numerical domain bruteforce for clusters. i.e.: www2, www3
* pickup new subdomains via reverse lookups (PTR records)
* better input validation
* improve function that validates target domain
if [[ $# -ne 1 && $# -ne 2 ]]
echo "usage: `basename $0` <domains-file> [results-path]";
echo "e.g.:";
echo "`basename $0` domains.txt";
echo "`basename $0` domains.txt /tmp/";
for i in `cat $1`
if [[ $# -eq 1 ]]
dnsmap $i
elif [[ $# -eq 2 ]]
dnsmap $i -r $2
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
1 - read subdomains from built-in list
1.1 - without saving output
1.2 - saving output
1.2.1 - txt format
1.2.2 - CSV format
2 - read subdomains from wordlist file
2.1 - without saving output
2.2 - saving output
2.2.1 - txt format
2.2.2 - CSV format
This diff is collapsed.
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment