Rainbow Crack Tutorial

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 6

RainbowCrack Tutorial

Introduction
RainbowCrack is a general propose implementation of Philippe Oechslin's faster time-memory trade-off technique. Function of this software is to crack hash. The straightforward way to crack hash is brute force. In brute force approach, all candidate plaintexts and corresponding hashes are computed one by one. The computed hashes are compared with the target hash. If one of them matches, the plaintext is found. Otherwise the process continues until finish searching all candidate plaintexts. In time-memory tradeoff approach, the task of hash computing is done in advance with the results stored in files called "rainbow table". After that, hashes can be looked up from the rainbow tables whenever needed. The pre-computation process needs several times the effort of full key space brute force. But once the one time pre-computation is complete, the table lookup performance can be hundreds or thousands times faster than brute force. This document explains the steps to make the RainbowCrack software working for first time user. Most contents in this document are implementation specific, while others are generic to time-memory tradeoff algorithm. The RainbowCrack software includes three tools that must be used in sequence to make things working. Step 1: Use rtgen program to generate rainbow tables. Step 2: Use rtsort program to sort rainbow tables generated by rtgen. Step 3: Use rcrack program to lookup rainbow tables sorted by rtsort. The table lookup process in final step is equivalent to the hash cracking process. The way to use these programs will be explained in this document. All of them are command line programs.

Step 1: Use rtgen program to generate rainbow tables


The rtgen program need several parameters to generate a rainbow table, the syntax of the command line is: rtgen hash_algorithm charset plaintext_len_min plaintext_len_max table_index chain_len chain_num part_index Explanation of these parameters: parameter meaning hash_algorithm The hash algorithm (lm, ntlm, md5 and so on) used in the rainbow table. The charset of all plaintexts in the rainbow table. All possible charset are charset defined in the charset.txt file. These two parameters define the possible length of all plaintexts in the plaintext_len_min rainbow table. If charset is numeric, plaintext_len_min is 1, and plaintext_len_max plaintext_len_max is 5. Then the plaintext "12345" is likely included in the table, but "123456" will not be included. These four parameters are really difficult to explain in simple words. To read and understand Philippe Oechslin's original paper can help to know the exact meaning. The table_index is related to the "reduce function" that is used in rainbow table. table_index The chain_len is the length of each "rainbow chain" in the rainbow table. A chain_len "rainbow chain" sized 16 bytes is the smallest unit in a rainbow table. A chain_num rainbow table contains lots of rainbow chains. part_index The chain_num is the number of rainbow chains in the rainbow table. The part_index parameter determines how the "start point" in each rainbow chain is generated. It must be a number (or begin with a number) in RainbowCrack 1.3 & 1.4. In RainbowCrack 1.2, this parameter can be any string because random "start point" is used, while 1.3 & 1.4 use the sequential "start point". The right values of all the parameters depend on what you need, to select good parameters require some understanding of the time-memory tradeoff algorithm. One ready to work configuration is given below, as an example: hash_algorithm lm, ntlm or md5 alpha-numeric = [ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789] charset or loweralpha-numeric = [abcdefghijklmnopqrstuvwxyz0123456789] plaintext_len_min 1

plaintext_len_max 7 chain_len 3800 chain_num 33554432 36^1 + 36^2 + 36^3 + 36^4 + 36^5 + 36^6 + 36^7 = 80603140212 key space table size key space is the number of possible plaintexts for the charset, plaintext_len_min and plaintext_len_max selected. 3 GB 0.999 The time-memory tradeoff algorithm is a probabilistic algorithm. Whatever the parameters are selected, there is always probability that the plaintext within the selected charset and plaintext length range is not covered. The success rate is 99.9% with the parameters used in this example. The actual rtgen commands used to generate the rainbow tables are: rtgen md5 loweralpha-numeric 1 7 0 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 1 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 2 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 3 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 4 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 5 3800 33554432 0 If ntlm or lm table is desired, replace "md5" in commands above with "ntlm" or "lm". If alpha-numeric charset is desired, replace "loweralpha-numeric" in commands above with "alpha-numeric". If lm table is to be generated, please CONFIRM the charset is alpha-numeric instead of loweralpha-numeric. The lm algorithm NEVER uses lowercase letters as plaintext. Now it is time to generate rainbow table. Change the current directory of your command prompt to RainbowCrack's directory, and execute following command: rtgen md5 loweralpha-numeric 1 7 0 3800 33554432 0 This command takes about 4 hours to complete on Core2 Duo E7300 processor. It is safe to stop the computation any time by pressing Ctrl+C. Next time if the rtgen program is executed with exactly same command line parameters, it will resume from where the computation is stopped and continue the table generation. When the command is finished, a file named "md5_loweralpha-numeric#1-

success rate

table generation commands

7_0_3800x33554432_0.rt" sized 512 MB will be in place. The file name is simply all the command line parameters connected, with the "rt" extension. The rcrack program to be explained later need this piece of information to know parameters of the rainbow table. So don't rename the file. Remaining tables can be generated in same way with commands: rtgen md5 loweralpha-numeric 1 7 1 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 2 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 3 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 4 3800 33554432 0 rtgen md5 loweralpha-numeric 1 7 5 3800 33554432 0 Finally, these files are generated: md5_loweralpha-numeric#1-7_0_3800x33554432_0.rt md5_loweralpha-numeric#1-7_1_3800x33554432_0.rt md5_loweralpha-numeric#1-7_2_3800x33554432_0.rt md5_loweralpha-numeric#1-7_3_3800x33554432_0.rt md5_loweralpha-numeric#1-7_4_3800x33554432_0.rt md5_loweralpha-numeric#1-7_5_3800x33554432_0.rt Now the rainbow table generation process complete.

512MB 512MB 512MB 512MB 512MB 512MB

Step 2: Use rtsort program to sort rainbow tables


The rainbow tables generated by rtgen program need some post processing to make table lookup easier. The rtsort program is used to sort the "end point" of all rainbow chains in a rainbow table. Use following commands: rtsort md5_loweralpha-numeric#1-7_0_3800x33554432_0.rt rtsort md5_loweralpha-numeric#1-7_1_3800x33554432_0.rt rtsort md5_loweralpha-numeric#1-7_2_3800x33554432_0.rt rtsort md5_loweralpha-numeric#1-7_3_3800x33554432_0.rt rtsort md5_loweralpha-numeric#1-7_4_3800x33554432_0.rt rtsort md5_loweralpha-numeric#1-7_5_3800x33554432_0.rt Each command above takes about 1 to 2 minutes to complete. The rtsort program will write the sorted rainbow table to the original file. Don't interrupt the rtsort program; otherwise the rainbow table being sorted will be damaged.

If the free memory size of your system is smaller than the size of the rainbow table being sorted, temporary hard disk space as large as the rainbow table size will be needed to store intermediate data. Now the rainbow table sorting process complete.

Step 3: Use rcrack program to lookup rainbow tables


The rcrack program is used to lookup the rainbow tables. It only accepts sorted rainbow tables. Assume the sorted rainbow tables are placed in c:\rt directory, to crack single hash the command line will be: rcrack c:\rt\*.rt -h your_hash_comes_here The first parameter specifies the path to the rainbow tables to lookup. The "*" and "?" character can be used to specify multiple files. Normally it takes seconds or tens of seconds to finish, if the plaintext is within the selected charset and plaintext length range. Otherwise, it takes much longer time to search all the tables only to find nothing. To crack multiple hashes, place all the hashes in a text file with each hash in a line. And then specify file name in rcrack command line: rcrack c:\rt\*.rt -l hash_list_file If the rainbow tables you generate use lm algorithm, the rcrack program has special support for it with the "-f" command switch. A hash dump file in pwdump format is required as input to rcrack program. The file will looks like this: Administrator:500:1c3a2b6d939a1021aad3b435b51404ee:e24106942bf3 8bcf57a6a4b29016eff6::: Guest:501:a296c9e4267e9ba9aad3b435b51404ee:9d978dda95e5185bbe da9b3ae00f84b4::: The pwdump file is the output of pwdump2, pwdump3 or other utilities. It contains both the lm hash and the ntlm hash. To crack lm hashes in pwdump file, use following command:

rcrack c:\rt\*.rt -f pwdump_file The lm hash algorithm converts all lowercase letters in plaintext to uppercase; as a result all the plaintexts cracked via the lm hash never contain lowercase letters, while the actual plaintext may contain lowercase letters. The rcrack program will try to do case correction with the ntlm hashes stored in same file and output the original plaintext.

You might also like