FAQ

	John the Ripper FAQ.

The latest version of this FAQ may be viewed online at:

	https://www.openwall.com/john/doc/FAQ.shtml


	Help!  I can't run John.

If you're not familiar with your OS, you should probably not be using
John in the first place since John is primarily a tool for system
administrators.  This is starting to change with the "community
enhanced" -jumbo versions' support for things such as password-protected
archives, though.

Here are the answers to a few (not very) common questions to avoid
having them asked over and over and for amusement.  For more serious
matters, please skip over to the next section.

Q: When I type "john" (or "john passwd", etc.), it says "command not
found" (or equivalent)?!
A: The examples given in John the Ripper documentation assume that you
know how to invoke newly-built programs from your shell.  On Unix-like
systems, it is typical to not have "." (the current directory) in your
$PATH (the list of directories to search for programs).  In that case,
you need to type "./john" (dot, slash, and "john", without the quotes)
to invoke the John binary executable located in the current directory.

Q: ...but I am on a Unix-like system and I don't seem to readily have a
John binary executable.
A: Please follow the instructions in INSTALL.

Q: When I double-click on "john.exe", a window flashes and disappears?!
A: You're not supposed to click.  You're supposed to run John from a
command-line shell.  On Windows, some of those shells would be cmd.exe,
command.com, or bash (the latter is available with Cygwin).


	Other typical new user questions.

Q: How do I start John on my password file, use a specific cracking
mode, see the passwords it cracked, etc?
A: See README and EXAMPLES. :-)

Q: Why doesn't John load my password file?  It says "No password hashes
loaded", "No password hashes loaded (see FAQ)", or "No password hashes
left to crack (see FAQ)".
A: Your password file taken from a Unix-like system might be shadowed.
You need to get both /etc/passwd and the shadow file (typically
/etc/shadow or /etc/master.passwd), and combine them into one file using
"unshadow" (which is supplied with John).  Please refer to EXAMPLES.
A: All of the password hashes found in the file (that are of the same
type as the very first recognized hash in the file unless you're using
the "--format=..." option) might be already cracked by previous
invocations of John.  (The message printed in that case has been changed
to "No password hashes left to crack (see FAQ)" starting with version
1.7.7.)  To display cracked passwords, use "john --show" on your
password hash file(s).  To force John to crack those same hashes again,
remove the john.pot file.
A: With PWDUMP-format files, John focuses on LM rather than NTLM hashes
by default, and it might not load any hashes at all if there are no LM
hashes to crack.  To have JtR Pro or a -jumbo version focus on NTLM
hashes instead, you need to pass the "--format=nt" option.  You'll also
need to use this option along with "--show".
A: If you're using the "--format" option, try dropping it.  Except for
the special case mentioned in the answer above, "--format" is normally a
way to choose one of multiple hash/cipher types found in the same file
or to clarify the hash/cipher type if it would otherwise be ambiguous
(e.g., a 32 hexadecimal character string may correspond to one of many
distinct hash types).  That is, you normally only need to use "--format"
when John would otherwise misdetect your hash/cipher type (e.g., when it
says LM and you know that your hashes are in fact raw MD5, you'd use
"--format=raw-md5" with -jumbo) or if it would load undesired entries
from the file.  If John does not load anything, then your use of
"--format" is probably unreasonable (or you should be using a different
version/build of John - see the answer below).
A: Your password hash or cipher type(s) might not be supported by John,
or at least by the version and build of John that you're using.  If
you're using a non-jumbo version, you will likely want to try -jumbo
instead, which supports a lot of additional hash and cipher types (e.g.,
you currently need -jumbo for raw MD5).  If unsuccessful with that and
if other answers (above and below this one) don't apply, please post a
note to the mailing list (see CONTACT) including a sample password file
line that John does not load (please make sure that the password is
already changed by the time you post).
A: John only loads properly formatted text files directly.  It can load
/etc/passwd and PWDUMP format files.  Starting with version 1.7.6, it
can also load text files containing one password hash per line (and
nothing else on that line).  Some other file formats are supported via
extra tools (supplied with John): unafs (Kerberos AFS database files),
undrop (Eggdrop IRC bot userfiles), ssh2sshng.py (OpenSSH private keys),
pdf2john (some password-protected PDF files), rar2john (some
password-protected RAR archives), zip2john (some password-protected
PKZIP and WinZip archives).  You need -jumbo for most of these.  To use
the proper one of these (for your file format), run it on your file(s)
and redirect the output to a new file (using your shell's output
redirection feature - e.g., "./ssh2sshng.py ~/.ssh/id_rsa > sshpasswd").
Then run John on the resulting file (e.g., "./john sshpasswd").
A: The file you're trying to run John on might in fact not be a password
file at all.
A: Your command line syntax might be wrong, resulting in John trying to
load a wrong file.

Q: John appears to misdetect my hash type.  I have raw MD5 hashes from a
web application, but John wrongly says they're LM hashes.  How do I get
them detected correctly?
A: Some hash and cipher types use ambiguous encodings - e.g., a 32
hexadecimal character string may correspond to one of many hash types,
including raw MD5, LM, NTLM, and many others supported in -jumbo.  First
of all, you need a version and build of John that supports your hash and
cipher type.  Starting with version 1.7.7 (and 1.7.7-jumbo*) John will
suggest alternate hash and cipher types for encodings that it finds
ambiguous (that is, those corresponding to more than one of its
supported hash and cipher types).  When doing so, it will suggest
specific "--format=..." options to use.  For example, when you run a
recent enough -jumbo version on raw MD5 hashes, it loads those as LM
(because they could in fact be LM, as well as for compatibility with
non-jumbo), but it suggests that you use "--format=raw-md5", which is
what you should in fact use in this case.  It makes other suggestions as
well because it does not know whether your hashes are raw MD5 or
something else.  You're supposed to know this and choose the right one
of the suggested "--format=..." options.  If you're not getting a
suggestion like this from John 1.7.7 or newer even though you're not yet
using the "--format" option, this means that your version and build of
John does not recognize the encodings as ambiguous, which may mean that
it does not support the actual hash or cipher type that you have in
mind.  If you're already using the "--format" option, try dropping the
option to receive the suggestions.  If you're using a non-jumbo version
of John, the first step is for you to try -jumbo instead.  As of this
writing, you do need -jumbo for some popular hash types such as raw MD5
and NTLM.

Q: What do the various numbers printed on the status line mean?
A: As of version 1.8.0, the status line may include: successful guess
count ("g"), session duration (in the D:HH:MM:SS format for days, hours,
minutes, and seconds), progress indicator (percent done and optionally
pass number out of the total number of passes), up to four speed metrics
("g/s", "p/s", "c/s", and "C/s"), and the current (range of) candidate
password(s) being tested (John is often able to test multiple candidate
passwords in parallel for better performance, hence a range).  The four
speed metrics are as follows: g/s is successful guesses per second (so
it'll stay at 0 until at least one password is cracked), p/s is
candidate passwords tested per second, c/s is "crypts" (password hash or
cipher computations) per second, and C/s is combinations of candidate
password and target hash per second.  Versions of John prior to 1.8.0
displayed only the C/s rate (calling it c/s).  When you restore a
pre-1.8.0 session with version 1.8.0 or newer, only the g/s and C/s
rates will be displayed, because the older .rec file format lacked
information needed to compute p/s and c/s.

Q: I am running John for 10 days and it is still not finished?!
Q: How long should I expect John to run?
A: It primarily depends on the cracking mode(s) and on your password
files (in particular, the type of hashes and the number of different
salts, if applicable).  Most importantly, you should note that the
"incremental" mode, which a default John run (with no command line
options) proceeds with after being done with the quicker checks, is not
supposed to terminate in a reasonable time.  It is up to you to decide
how long you're going to let it run, then consider any uncracked
passwords strong enough.  "Single crack" mode runs typically take from
under a second to one day (depending on the type and number of password
hashes).  Wordlist mode runs may also be quick (under a second) for
tiny wordlists and fast hashes or they may take multiple days with large
wordlists, with word mangling rules, and with slow hash types and
substantial numbers of different salts.  The status line John reports
whenever you hit a key includes a progress indicator (percent complete)
for "single crack" and wordlist modes.  With no cracking mode requested
explicitly, John will start with "single crack" mode (pass 1), then
proceed with wordlist mode (pass 2), and finally with "incremental" mode
(pass 3).  The pass numbers are reported on the status line, too.  It is
reasonable to let John reach "incremental" mode (pass 3) and run that
for a while (some days).  You will notice that John's success rate (the
number of passwords cracked per hour or per day) will be dropping
rapidly.  When you determine that the success rate is low enough, you
interrupt John.

Q: Does John support multi-processing or distributed processing?
A: Yes, but you need to explicitly enable this if desired.  Starting
with version 1.8.0, there's the "--fork" option on Unix-like systems (to
make use of multiple CPUs and/or CPU cores in a single system) and the
"--node" option on all systems (this one allows for a trivial form of
distributed processing).  The "--fork" and "--node" options may also be
used together.  Please refer to OPTIONS for a description of these
options.  Additionally, there's built-in parallel processing support
using OpenMP for all crypt(3) hash flavors (DES-, MD5-, and
Blowfish-based) supported by John natively, and when running on Linux or
Solaris also for the underlying system's thread-safe password hashing
function.  The latter is only reasonable to use for crypt(3) hash types
not yet supported by John natively (such as for glibc 2.7+ SHA-crypt
hashes as used by recent versions of Fedora and Ubuntu, and for SunMD5
hashes, which may optionally be enabled on Solaris).  In "community
enhanced" -jumbo versions, parallelization with OpenMP is also supported
for many (but not all) of the hash and cipher types added in those
versions (including for their built-in implementation of SHA-crypt).
To use John's OpenMP support, you need to either use an existing
OpenMP-enabled build (e.g., "john-omp.exe" on Windows) or make an
OpenMP-enabled build by uncommenting one of the OMPFLAGS lines near the
beginning of Makefile.  This requires GCC 4.2 or newer, or another
OpenMP-capable C compiler.  For other hash or cipher types and/or to
distribute the workload between multiple machines, other approaches need
to be used.  One of those approaches is to use the "--fork" and "--node"
options.  For a very small number of nodes (CPUs, CPU cores, and/or
machines), it is also reasonable to use a manual approach, such as to
have your nodes try different password lengths.  This is easily
accomplished with "incremental" mode's "MinLen" and "MaxLen" settings
(see CONFIG).  You might not need to split the workload for "single
crack" and wordlist modes since these are typically relatively quick,
although "--fork" and "--node" are supported for these modes too.  You
may safely run multiple instances of John in the same working directory,
all writing to the same "pot file" (this is a feature).  You do,
however, need to assign each of them a unique session name, with
"--session" (please note that doing so does not eliminate the need to
also distribute the workload with "--node" or otherwise, as discussed
above).  Other approaches, such as splitting password files naively
(without regard to salts), are typically less efficient (in some cases
to the extent where there's no speedup from using multiple nodes at
all).  Some other approaches, such as using MPI, are listed on the wiki
at: https://openwall.info/wiki/john/parallelization

Q: Where do I get wordlists for use with John?
A: https://www.openwall.com/wordlists/

Q: Where do I get new versions of John the Ripper?
Q: Where do I get the source code for John?
Q: I only have the source code for John the Ripper, where do I get it
pre-compiled for my OS (if supported)?
Q: What is the primary website for John the Ripper?
A: https://www.openwall.com/john/

Q: How can I contact you (the author)?
A: See CONTACT.


	Questions sometimes asked by existing users.

Q: I've recently switched my system to Blowfish-based password hashes,
but there are still some DES-based and MD5-based hashes in the password
file.  How do I handle multiple hash types in one file?
A: Use the "--format=..." option to tell John which hashes you would
like it to load.  Unfortunately, you will have to run John for each hash
type separately.  This requirement may sometimes be avoided with the use
of "--format=crypt", but this is not recommended.  Please see the
description of the "--format" option in OPTIONS for more detail.

Q: I have 10 users, but John said it loaded 15 password hashes.  What's
going on?
A: Some extremely poorly designed hash types (Windows LM hashes and
DES-based crypt(3) hashes known as "bigcrypt") have a property that
allows John to split their encodings into two separate hashes
(corresponding to "halves" of plaintext passwords) on load.  John then
proceeds to crack those hashes separately, so at a given time it might
have only one of two halves of some passwords cracked.  If interrupted
and restarted, it would need to only load the hashes that correspond to
uncracked password halves, so the number of such hashes is what John
reports (in all cases, for consistency).

Q: Are the strings tried with "-i" ("incremental" mode) random?  They
certainly look like they are almost random.
A: No, they are not.  No single candidate password will be tried for a
second time and the order in which they are tried is in fact very smart:
it is based on frequencies of different trigraphs, stored and processed
separately for each character position and for each password length.

Q: Why doesn't John display a progress indicator for the "incremental"
mode?
A: Do you really want to see a 0% all the time?  As explained in MODES,
"incremental" mode is not supposed to terminate in a reasonable time.
(There are a few exceptions to this, so a progress indicator has been
added in -jumbo and it might be added in official versions later.)

Q: I just noticed that the p/s, c/s, and C/s rates reported while using
"incremental" mode are a lot lower than they are with other cracking
modes.  Why is that?
A: You're probably running John for a few seconds only.  The current
"incremental" mode implementation uses large character sets, which need
to be expanded into even larger data structures in memory each time John
switches to a different password length.  Fortunately, this is only
noticeable when John has just started since the length switches become
rare after a few minutes.  For long-living sessions, which is where we
care about performance the most, this overhead is negligible.  This is a
very low price for the better order of candidate passwords tried.

Q: What are the "real" and "virtual" c/s rates as reported by "--test"?
A: These correspond to real and virtual (processor) time, respectively.
When running single-threaded, the two results are normally almost the
same, but the "real" c/s rate becomes smaller when the system is under
other load, with the "virtual" c/s rate indicating roughly what you
could expect to get from the same system if it were not loaded.  When
running multi-threaded, the "real" c/s rate is normally much higher than
the "virtual" c/s rate, with the latter roughly indicating performance
of one thread on an otherwise idle system.

Q: How can I test John's password hashing routines for proper operation?
A: John always performs a self-test when you run it on a password file
and refuses to work if an error occurs.  If you need to test all of the
low-level routines at once, use "--test".

Q: What is the format of the crash recovery files ("john.rec", other
.rec's)?  What do the numbers mean?
A: The format of these files is deliberately undocumented and is subject
to change without notice.  (However, each release of John the Ripper is
likely to be able to read .rec files produced by at least the
immediately preceding release.  Whenever compatibility is broken, John
will refuse to recover the session, leaving the .rec file intact.)
Although the meaning of some of the numbers that get into .rec files is
trivial to explain, it is not possible to reasonably describe some
others without going into great detail on John internals.  If you really
need to know, read the source code.