Info Harvesting

Docs that have proven to be a staple in understanding computer/network security. This is not an inclusive forum and nothing ipublished will tell you how to 0wn someone, these docs will help you understand how you got 0wnd.
Post Reply
User avatar
Posts: 1688
Joined: Sun Jul 07, 2002 10:02 am
Location: any given

Info Harvesting

Post by weazy » Fri May 30, 2003 7:27 pm

I get a lot of emails from people asking me how they can break into their ISP or how they can break into a system etc etc. Infact, such questions are almost the most common ones, from all the questions I get. Well, after this popular demand, I thought that an entire manual on breaking into systems was needed. So here goes..


You see, breaking into systems or getting root on a system is not as difficult as it seems. And it by no means requires you to be an Uberhacker. Getting into a system is quite easy and it requires you to know at least one programming language (preferably C), and have a more than an average IQ.  However, breaking into systems does require a bit of luck and also a bit of carelessness or stupidity on the part of the system administrator of the target system.


What I mean to say by all this is that, breaking into systems is no big deal, anyone could do that, even a script kiddie, however, the part of the entire Hacking process where more than most people falter is the remaining undetected part. Anonymity or remaining anonymous to the Server logs and preventing detection of a break-in is the most difficult part of Hacking into a system.


What separates a good Hacker from a Script Kiddie or a Lamer is that the former has more than several ways of making sure that no one even suspects that there has been a break in, while on the other hand, the later has no clue what so ever as to what he is doing or what he needs to do to prevent such detection. There are so many ready to Use canned C programs or Hacking utilities available on the net, that a huge number of wannabe hackers, download them and use them to Hack into systems. Well, not only do they do not work properly and flawlessly, they also provide no mechanism of remaining anonymous. What is more, say if you are not using a canned Hacking tool, and are also not trying to remain anonymous, then you stand a greater chance of remaining undetected than if you were using such a tool. So think before you use such tools, you might be able to get the Password file and become very kewl, however, you will certainly be caught later if not sooner.


The first step that you need to take once you have decided the target computer is to find out as much information as you can about it. You see, to break into a system you need to exploit a vulnerability existing in the services offered by it. Almost all systems have certain open ports, which have certain daemons or services running on them.



HACKING TRUTH: There are two types of ports. There are hardware ports, which are the slots existing behind the CPU cabinet of your system, into which you plug-in or connect your hardware to. For Example, COM1, COM2, Parallel Port etc.  However, we are not interested in such ports. We are concerned with the other type of ports, which are the virtual or the software ports. Such a virtual port is basically a virtual pipe through which information goes in and out. And all open ports have a service or daemon running on it. A service or a daemon is nothing but the software running on these ports, which provide a certain service to the users who connect to it. For Example, Port 25 is always open on a server handling mails, as it is port where the Sendmail service is running by default.



So basically the first step in your quest to breaking into a system is to get as much information on it, as you can. Try to get, the list of open ports, the list of services running on the respective open ports and whole lots of other kind of information to which I will come later.


Anyway, so firstly, get a good Port Scanner, preferably stealth and then do a port scan on the target host. Now one thing that you must remember while doing a port scan is the fact that there are various so called ‘stealth’ port scanners around which claim to be undetectable, however most of them are detectable. So instead of using such’ false claims’ port scanners, I suggest you code one on your own.


But why do I need to use a stealth Port Scanner and how can I code my own Port Scanner? Well, the reason as to why you need a stealth port Scanner is that many system administrators log all port scans and records the IP and other information on such attempts, this makes you susceptible to getting caught. In my opinion the best Port Scanners around are those, which send SYN/FIN packets from a spoofed host, making logging useless. Such a port Scanner would be coded in C, but will not run in Windows. This was just an idea, now it is up to you to code it yourself.


Anyway, let me assume that you have got hold of a good ‘impossible to detect’ Port Scanner, now scan the target system for all open ports and record the open lists:


Note: In this manual, I have taken up my ISP as an example target system. It would be foo-barred throughout as


In my case, I found that the following ports were open:


Port Number                Service


21                            FTP        

23                            Telnet

25                            SMTP

53                            DNS

79                            Finger

80                            HTTP    

110                          POP

111                          Not Useful

389                          Not Useful

512                          rlogin


Note: Only a few Port Scanners give you both the open Ports and the services running on them. Most Port Scanners only return the list of Open Ports. This is fine too; as once you get the list of open ports then you can find out the corresponding services running on them, referring to the RFC 1700. It contains the complete list of Port Numbers and the corresponding popularly running services.


Now port scanning takes advantage the 3-stage TCP handshake to determine what ports are open on the

remote computer. To learn more about the TCP\IP protocol read the networking manuals that I distribute on

my mailing list.


Tools like SATAN and lots of them more allow you to find out the list of open ports, the daemon or the service running at each open port and also the service's vulnerability at the click of a button. You can't call yourself a hacker if you need some Software, which first of all is not written by you to do something as lame as a port scan. Well yes I do agree that looking for open ports manually on a server would take a long time. But what I am suggesting is that you use a Port Scanning tool, which just gives you a list of open ports without the list of services and the vulnerabilities. I assure you, if you try and explore an open port of a remote server manually, you will be able to learn more about the remote system and also it will give you a taste of what hacking actually is. If you use a port scanner, which gives you all details at the click of a button to impress your friends, let me assure you none of them will be impressed, as I am sure anyone can use SATAN and other such scanners.


Another thing you need to be careful about before port scanning your ISP is that most port scanners are very easily detected and can easily be traced and you have no excuse if you are caught doing a port scan on a host., it a sure sign of Hacker Activity. There are many stealth scanners like Nmap, which claim to be untraceable. But the truth is that they are very much traceable and they are quite inaccurate as they send only a single packet to check if a port is open or not. And if the host is running the right kind of Sniffer software maybe Etherpeek then the Port scan can be easily detected and the IP of the user logged. Anyway some ISP's are really afraid of Hacking activities and even at the slightest hint of some suspicious hacking activity something like Port scanning, they can disable your account. So just be careful.


Evil Hacking Trick: Well try to keep an eye on TCP port 12345, and UDP port 31337 these are the default

ports for the popular trojans NetBus and BO, respectively



Some ISP's are quite aware of Hacking Activities and are one step ahead. They may be running some excellent software, which will keep hackers away. EtherPeek is an excellent example of sniffing software, which can easily trace users who are port scanning. Nuke Nabber a Windows freeware claims to be able to block Port Scans. I have not tested it so I can't say for sure. Then there is another fun program known as

Port Dumper, which can fake daemon (services) like Telnet, Finger etc. There is also some software, which will show a weird list of open ports. What I mean by that is, if you port scan a host running such software, then it will keep showing random open ports, and you port Scanning Software will go crazy.


Anyway, so once you get a list of open ports, start analyzing the weak points or the services which might help us to get more information about the target system which would prove invaluable to the breaking in process. Try to exploit the commands or the options available on each open port to either find a vulnerability, which could be exploited, or some kind of information on the target system. That is pretty much the only kind of things that we would be looking for. Now, let me explain how I try to find out such things with the list of open ports (of my ISP) and services running on them.


Note: Before proceeding, refer to the table of open ports which we got earlier(of host in the manual and yes, I am starting from Port 79 as if I start from Port 21, then the manual will become very very short.


It has Port 79 open or in other words, has finger running, however, almost all Finger daemon are configured to not return much information about Users, however, let us try some common Finger exploits which can sometimes very very rarely get you root.


finger root

finger system



These exploits are very very old and do not work almost 99 times out of 100. So the Finger port is ruled out.


Now let us move on, in the list of open ports, the HTTP port or Port 80 is also open, this means that this target system probably maintains a web site. So let me launch my favorite browser (Internet Explorer, if you are interested.) and see what they have on their site. Well, actually we are not even remotely interested in what they have on their site, but what we are interested in is to see, whether they have the CGI-BIN directory open to public or not, an dif yes if any of the common CGI exploits, which get you root, work or not.


So I type in the following in the URL box of my browser:


I get the NOT FOUND error message; this probably means that this system does not support CGI-Scripts. If the CGI-Bin directory had been blocked from public access, then we would probably have gotten the Forbidden Error Message.


However, finding out that our target system does have the CGI-Bin directory cannot be said to be disappointing as the known CGI exploits are almost primitive and finding out new exploits should be kept out of this manual.


OK, so Port 80 and Port 79 are ruled out, they neither have any vulnerability nor do they give any information about the target system. [Well actually the HTTP port does give us some valuable information, but we will come to that later.]


Anyway, so let us try Port 21 or the FTP port. Now, there are two ways of connecting to Port 21 of a host, the first one is to telnet to Port 21 and other one is to use the MS-DOS FTP client. You could choose any of the two for this section, however, I kind of like the command line FTP client, although many people say it is lame. Anyway, so I launch up a FTP connection to




Connected to



220-#           Welcome to MTNL's ftp site



220-#  You can upload your own homepages at this site!!!


220-#  Just login with your username and upload the HTML pages.

220-#  (You can use your favourite HTML editor as well)


220-#  World will see it at


220-#  So get going......UNLEASH YOUR CREATIVITY !!!!




220 FTP server ready.


We are immediately greeted by the FTP daemon banner, which tells us that this is the FTP server where, people using MTNL’s (My ISP) Internet Services, can upload their site. Now, normally FTP daemon banners are more informative than this one. They usually do give away the name of the Operating System running and also the FTP daemon running. Well, actually it is the login prompt of the daemon banner which gives us the Operating System running on it.  Normally, a typical daemon banner, would have the following Login prompt:


220 FTP server (Digital UNIX Version 5.60) ready.

User (


Notice the System name in the brackets on the first line. However, normally almost all FTP daemons are better configured (that is the case in the example target system: their login prompt is somewhat like the below:


220 FTP server ready.

User (


See, no Operating System name. However, with the help of some kewl commands, such systems too can be reveal the OS running on them. However, before we go on, there is one thing that you have to be clear about. Now, we had FTP’ed to, so you normally expect to connect to Port 21 of, however that is not true. (Atleast in this case.) If you look at the daemon banner again, then you would notice that the last line says:


220 FTP server ready.


Now how did that happen? Well, is Port 21 not open on ? Well, no and yes. What actually happens is that, Port 21 of is open and a daemon there is listening for connections. As soon as a connection is established, it transfers the control or connected the visitor to, which is on the same network as Now this, system is solely a FTP machine. It has no other services running.  So whatever information, we gather from such a FTP port is not of but of  Get it?


Anyway, when you get the login prompt, then login anonymously with the anonymous as the Username and a false email address as the password. 


220 FTP server ready.

User ( anonymous

331 Guest login ok, send your complete e-mail address as password.


230 User anonymous logged in.  Access restrictions apply.


Even if you have an account at the FTP server into which you plan to break in, it is always better not to use that pair of Username and Password. Logging in anonymously has many advantages. Say if you did cause some harm to the target system and if you use your (Nonanonymous) Username and Password pair, then if you were not able to edit the server logs you could get into some serious trouble. [Well actually not much, only say your account might be disabled.  However, it could be worse.]


Ok, you are in, now let us get the FTP client to tell us which commands are available by typing the help command.


ftp> help

Commands may be abbreviated.  Commands are:


!                 delete            literal            prompt

?               debug            ls                   put

append     dir                 mdelete         pwd

ascii          disconnect    mdir             quit

bell           get                mget             quote

binary       glob              mkdir           recv

bye            hash             mls               remotehelp

cd              help             mput             rename

close         lcd               open              rmdir


Uhmmm.,..,  none of the above commands seem to be or sound to be of use to us. So the ‘help’ command did not reveal any useful commands. However, you see the above list of commands are commands which are offered by the FTP client and almost more often than not, the FTP daemon offers a wider array of commands. To get a complete list of commands offered by the FTP daemon, use the ‘remotehelp’ command:


ftp> remotehelp

214-The following commands are recognized (* =>'s unimplemented).

   USER    PORT    STOR    MSAM*   RNTO    NLST    MKD     CDUP


   ACCT    TYPE    MLFL*   MRCP*   DELE    SYST    RMD     STOU

   SMNT*   STRU    MAIL*   ALLO    CWD     STAT    XRMD    SIZE

   REIN*   MODE    MSND*   REST    XCWD    HELP    PWD     MDTM


214 End of help


Note: To get a single line description of each command, type help followed by a space and the command of which you want a description.


One thing to remember here is that to execute any command from the remote FTP commands list you need to make use of the ‘literal’ keyword. What  I mean by that is that all remote FTP commands have to be preceded by the word ‘literal’. For example, say you want to execute the remote FTP command: ‘stat’, then you would type:


ftp> literal stat


HACKING TRUTH: According to FTP help, the literal command is described as:

ftp> help literal

literal         send arbitrary ftp command



Anyway, amongst the remote FTP commands, the commands of interest to us are-: ‘stat’ and ‘syst’. Let us see what they return when executed-:


ftp>literal stat

211- FTP server status:

     Version 5.60

     Connected to 203.xx.251.198 (203.xx.251.198)

     Logged in anonymously

     TYPE: ASCII, FORM: Nonprint; STRUcture: File; transfer MODE: Stream

   211- No data connection

211 End of status


Note: The IP address is of and not your machine.


ftp> literal syst

215 UNIX Type: L8 Version: BSD-198911


Voila, we get the Operating System name running on At last some useful information.


Finger and HTTP both failed, what do we do now? Let us turn to the den of the Buggiest daemon on Earth i.e. Sendmail: Port 25, the SMTP port.


Sendmail is certainly the buggiest daemon on earth; it has the highest number of known exploits amongst all the daemons. So this probably should get us through. Let us telnet to Port 25 and find out whether an exploitable version of Sendmail is running.


C:\windows> telnet 25


220 ESMTP Sendmail 8.9.1 ( Thu, 29 Jun 2000 14:18:12 0530 (IST)


When you telnet to Port 25, then the first thing that you come across would be a something like the above welcome daemon banner. A daemon banner is a Hacker’s best friend. It reveals important information about the host, which proves to be invaluable in breaking into it. It basically tells you which daemon or service is running on that port and also the version of that particular service. Like for example, in this case, the Sendmail daemon banner tells us that ESMTP Sendmail 8.9.1 is running and it also gives us other information about the host at which this service is running.


Anyway, getting back to the topic, this banner reveals a big vulnerability existing in the host computer. It tells us that is running an old, vulnerable version of Sendmail. The latest version is Sendmail 8.9.4 (correct me if I am wrong.), so this particular version of Sendmail wouldn’t be without any bugs.


So then what you do is visit PacketStorm or search at your favorite Hacking stuff related search engine for a C program which demonstrates how to exploit version 8.9.4 of Sendmail. Now, all this might sound a bit too simple, well it certainly isn’t, read on for more info.


Now, there are a couple of things that you need to keep in mind while getting this done. Say, you have found out that the victim runs Sendmail 8.9.4, now you cannot simply break in by running any exploit for this version. By that what, I mean to say is that, an exploit, which is coded to be executed on a Linux platform, will not work if you try to compile and run it on a Windows platform. So basically before you execute the ‘kewl’ exploit program that you downloaded, you should find out which platform it is meant for and if you are not running that platform, then you will need to get your gray cells working.


This is the stage where real hackers are differentiated between script kiddies, this is when those people who really know something prevail.  Normally say if a exploit is designed to work on Linux, then if you edit its code and change its header files (if necessary), then that particular exploit can be made to run on Windows too. However, there are certain exploits, which simply would not run on a different OS than it is designed too.


Anyway, let us get back to point. You have edited the exploit code and made it compatible with your platform. Now what else? Another thing that you want to keep in mind is the Operating System, which the exploit can exploit. You see, there are certain exploits, which work only if the victim system is running a specific Operating System. For Example,


There was once a Sendmail hole, which worked only if the target System was running Sun OS without which, it simply refused to even work.


So in some cases it becomes necessary, to find out the Operating System running at the target system. Although not all exploits require the target system to be running a specific system, but why take a chance. Right?


So basically you should be aware of the following things while getting a ready to use exploit-:


1.)      The Daemon name and version you are trying to exploit For Example, Sendmail 8.9.4

2.)      The Operating System at which it is designed to run. (If necessary)

3.)      The operating System it requires the target system to be running. (If necessary)


That brings us to as to how to find out the Operating System running at the target system? Well, the HTTP port holds the key. Simply, telnet to Port 80 of the target system.


C:\windows>telnet 80


Now, once you get the input prompt, then, type an invalid HTTP command. For Example, X or Iamgreat or abc etc. Just type anything as long as it is not a valid HTTP command. Then press enter twice.



Hacking Truth: After each HTTP command one has to press Enter Twice to send the command to the server or to bring about a response from a server. It is just how the HTTP protocol works.



On Port 80 of my example target system, I type simply ‘ankit’ and press enter twice. This is the kind of response I get:


HTTP/1.1 400 Bad Request

Server: Netscape-Enterprise/3.5.1


The server replies with the version of HTTP it is running (not so important), it gives us an error message and the error code associated with it(again not so important), but it also gives us the OS name and OS version, it is running. Wow!!! It gives hackers who want to break into their server the ultimate piece of information, which they require.

--The Devil is in the Details--

Post Reply