Hack Guide
Hack Guide
**********************************************************************
| the author hereby grants permission to reproduce, redistribute, |
| or include this file in your g-file section, electronic or print |
| newletter, or any other form of transmission that you choose, as |
| long as it is kept intact and whole, with no ommissions, delet- |
| ions, or changes. (c) the mentor- phoenix project productions |
| 1988,1989 xxx/xxx-xxxx |
**********************************************************************
contents
~~~~~~~~
this file will be divided into four parts:
part 1: what is hacking, a hacker's code of ethics, basic hacking safety
part 2: packet switching networks: telenet- how it works, how to use it,
outdials, network servers, private pads
part 3: identifying a computer, how to hack in, operating system
defaults
part 4: conclusion- final thoughts, books to read, boards to call,
acknowledgements
there are two ways to find interesting addresses to connect to. the first
and easiest way is to obtain a copy of the lod/h telenet directory from the
lod/h technical journal #4 or 2600 magazine. jester sluggo also put out a good
list of non-us addresses in phrack inc. newsletter issue 21. these files will
tell you the nua, whether it will accept collect calls or not, what type of
computer system it is (if known) and who it belongs to (also if known.)
the second method of locating interesting addresses is to scan for them
manually. on telenet, you do not have to enter the 03110 dnic to connect to a
telenet host. so if you saw that 031104120006140 had a vax on it you wanted to
look at, you could type @c 412 614 (0's can be ignored most of the time.)
if this node allows collect billed connections, it will say 412 614
connected and then you'll possibly get an identifying header or just a
username: prompt. if it doesn't allow collect connections, it will give you a
message such as 412 614 refused collect connection with some error codes out to
the right, and return you to the @ prompt.
there are two primary ways to get around the refused collect message. the
first is to use a network user id (nui) to connect. an nui is a username/pw
combination that acts like a charge account on telenet. to collect to node
412 614 with nui junk4248, password 525332, i'd type the following:
@c 412 614,junk4248,525332 <---- the 525332 will *not* be echoed to the
screen. the problem with nui's is that they're hard to come by unless you're
a good social engineer with a thorough knowledge of telenet (in which case
you probably aren't reading this section), or you have someone who can
provide you with them.
the second way to connect is to use a private pad, either through an x.25
pad or through something like netlink off of a prime computer (more on these
two below.)
the prefix in a telenet nua oftentimes (not always) refers to the phone area
code that the computer is located in (i.e. 713 xxx would be a computer in
houston, texas.) if there's a particular area you're interested in, (say,
new york city 914), you could begin by typing @c 914 001 <cr>. if it connects,
you make a note of it and go on to 914 002. you do this until you've found
some interesting systems to play with.
not all systems are on a simple xxx yyy address. some go out to four or
five digits (914 2354), and some have decimal or numeric extensions
(422 121a = 422 121.01). you have to play with them, and you never know what
you're going to find. to fully scan out a prefix would take ten million
attempts per prefix. for example, if i want to scan 512 completely, i'd have
to start with 512 00000.00 and go through 512 00000.99, then increment the
address by 1 and try 512 00001.00 through 512 00001.99. a lot of scanning.
there are plenty of neat computers to play with in a 3-digit scan, however,
so don't go berserk with the extensions.
sometimes you'll attempt to connect and it will just be sitting there after
one or two minutes. in this case, you want to abort the connect attempt by
sending a hard break (this varies with different term programs, on procomm,
it's alt-b), and then when you get the @ prompt back, type 'd' for disconnect.
if you connect to a computer and wish to disconnect, you can type <cr> @
<cr> and you it should say telenet and then give you the @ prompt. from there,
type d to disconnect or cont to re-connect and continue your session
uninterrupted.
phone scanning
~~~~~~~~~~~~~~
finally, there's the time-honored method of computer hunting that was made
famous among the non-hacker crowd by that oh-so-technically-accurate movie
wargames. you pick a three digit phone prefix in your area and dial every
number from 0000 --> 9999 in that prefix, making a note of all the carriers
you find. there is software available to do this for nearly every computer
in the world, so you don't have to do it by hand.
unix- there are dozens of different machines out there that run unix.
while some might argue it isn't the best operating system in the
world, it is certainly the most widely used. a unix system will
usually have a prompt like 'login:' in lower case. unix also
will give you unlimited shots at logging in (in most cases), and
there is usually no log kept of bad attempts.
common accounts/defaults: (note that some systems are case
sensitive, so use lower case as a general rule. also, many times
the accounts will be unpassworded, you'll just drop right in!)
root: root
admin: admin
sysadmin: sysadmin or admin
unix: unix
uucp: uucp
rje: rje
guest: guest
demo: demo
daemon: daemon
sysbin: sysbin
iris- iris stands for interactive real time information system. it orig-
inally ran on pdp-11's, but now runs on many other minis. you can
spot an iris by the 'welcome to "iris" r9.1.4 timesharing' banner,
and the account id? prompt. iris allows unlimited tries at hacking
in, and keeps no logs of bad attempts. i don't know any default
passwords, so just try the common ones from the password database
below.
common accounts:
manager
boss
software
demo
pdp8
pdp11
accounting
nos- nos stands for networking operating system, and runs on the cyber
computer made by control data corporation. nos identifies itself
quite readily, with a banner of 'welcome to the nos software
system. copyright control data 1978,1987'. the first prompt you
will get will be family:. just hit return here. then you'll get
a user name: prompt. usernames are typically 7 alpha-numerics
characters long, and are *extremely* site dependent. operator
accounts begin with a digit, such as 7etpdoc.
common accounts/defaults:
$system unknown
systemv unknown
decserver- this is not truly a computer system, but is a network server that
has many different machines available from it. a decserver will
say 'enter username>' when you first connect. this can be anything,
it doesn't matter, it's just an identifier. type 'c', as this is
the least conspicuous thing to enter. it will then present you
with a 'local>' prompt. from here, you type 'c <systemname>' to
connect to a system. to get a list of system names, type
'sh services' or 'sh nodes'. if you have any problems, online
help is available with the 'help' command. be sure and look for
services named 'modem' or 'dial' or something similar, these are
often outdial modems and can be useful!
the above are the main system types in use today. there are
hundreds of minor variants on the above, but this should be
enough to get you started.
unresponsive systems
~~~~~~~~~~~~~~~~~~~~
occasionally you will connect to a system that will do nothing but sit
there. this is a frustrating feeling, but a methodical approach to the system
will yield a response if you take your time. the following list will usually
make *something* happen.
1) change your parity, data length, and stop bits. a system that won't re-
spond at 8n1 may react at 7e1 or 8e2 or 7s2. if you don't have a term
program that will let you set parity to even, odd, space, mark, and none,
with data length of 7 or 8, and 1 or 2 stop bits, go out and buy one.
while having a good term program isn't absolutely necessary, it sure is
helpful.
2) change baud rates. again, if your term program will let you choose odd
baud rates such as 600 or 1100, you will occasionally be able to penetrate
some very interesting systems, as most systems that depend on a strange
baud rate seem to think that this is all the security they need...
3) send a series of <cr>'s.
4) send a hard break followed by a <cr>.
5) type a series of .'s (periods). the canadian network datapac responds
to this.
6) if you're getting garbage, hit an 'i'. tymnet responds to this, as does
a multilink ii.
7) begin sending control characters, starting with ^a --> ^z.
8) change terminal emulations. what your vt100 emulation thinks is garbage
may all of a sudden become crystal clear using adm-5 emulation. this also
relates to how good your term program is.
9) type login, hello, log, attach, connect, start, run, begin, logon, go,
join, help, and anything else you can think of.
10) if it's a dialin, call the numbers around it and see if a company
answers. if they do, try some social engineering.
password list
=============
******************************************************************************
references:
1) introduction to itapac by blade runner
telecom security bulletin #1
2) the ibm vm/cms operating system by lex luthor
the lod/h technical journal #2
3) hacking the iris operating system by the leftist
the lod/h technical journal #3
4) hacking cdc's cyber by phrozen ghost
phrack inc. newsletter #18
5) usenet comp.risks digest (various authors, various issues)
6) usenet unix.wizards forum (various authors)
7) usenet info-vax forum (various authors)
recommended reading:
1) hackers by steven levy
2) out of the inner circle by bill landreth
3) turing's man by j. david bolter
4) soul of a new machine by tracy kidder
5) neuromancer, count zero, mona lisa overdrive, and burning chrome, all
by william gibson
6) reality hackers magazine c/o high frontiers, p.o. box 40271, berkeley,
california, 94704, 415-995-2606
7) any of the phrack inc. newsletters & lod/h technical journals you can find.
acknowledgements:
thanks to my wife for putting up with me.
thanks to lone wolf for the rsts & tops assistance.
thanks to android pope for proofreading, suggestions, and beer.
thanks to the urvile/necron 99 for proofreading & cyber info.
thanks to eric bloodaxe for wading through all the trash.
thanks to the users of phoenix project for their contributions.
thanks to altos computer systems, munich, for the chat system.
thanks to the various security personel who were willing to talk to
me about how they operate.