The AFS File System

What is AFS?
AFS is an enterprise file system designed for use in a distributed environment on multiple computing platforms. AFS allows users on various types of computers to access the same file system. To a casual UNIX user, AFS disk space looks like a regular local disk; to Windows and Mac OS X users, it looks like a normal network drive. So with a single namespace and Kerberos authentication, AFS allows a user to log into any machine participating in the DCI and be presented their files and/or applications.
AFS is a distributed file system which allows users to share and access all the files stored in the file system. It is as easy as accessing files stored on their local file system. AFS was created by Carnegie Mellon University (CMU) and IBM in 1985. The picture below shows AFS' basic structure.
An AFS name space has a very similar "look and feel" as a traditional UNIX tree. All UNIX commands to traverse the tree work just as usual. AFS is a way to allow users to manage files just like they do in the UNIX file system.

The Andrew File System (AFS) is a distributed networked file system which uses a set of trusted servers to present a homogeneous, location-transparent file name space to all the client workstations. It was developed by Carnegie Mellon University as part of the Andrew Project. It is named after Andrew Carnegie and Andrew Mellon. Its primary use is in distributed computing


How does AFS work?
Like any network application, AFS has two components, a client and a server. The client component resides on each machine that wants to use AFS. This client asks the server for files stored in AFS and the server sends the file to the client over the network. The client then presents the file to the user as if it were local to the machine. When the user makes any changes in AFS space, such as creating a new file, saving a file or deleting files, the client sends the information to the server where updates occur. AFS speeds this process up by using disk caching. The AFS client keeps pieces of commonly used files on local disk. When the user asks the AFS client for a file, it checks to see if that file is in the local disk cache. If it is, the client then asks the server if the file it has in disk cache is up-to-date. If the server reports that the file is up to date, the client is able to present the client with the file that is stored in the local disk cache rather than transferring the file from the server.
Why Use AFS?
AFS file space is available from any AFS client installed on any supported platform. AFS provides the flexibility of having the same file structure and content available from any supported client, whether that client is Windows XP/2000/NT or a variety of UNIX clients.
Additionally, AFS file space is managed centrally by ATN, freeing departments from having to buy and maintain departmental file servers. It also provides end users with the ability to store data in a secure and reliable location.





AFS at UNC
ATN maintains an AFS cell, isis.unc.edu, which houses departmental and personal directories as well as shared applications. To be able to see this AFS space, you will need to use a machine that has an AFS client installed on it. If you would like to install an AFS client and your machine is supported, you can follow these instructions:
• AIX
• IRIX 6.5
• Linux 2.4 (RedHat)
• Linux 2.6 (RedHat)
• Mac OS X
• Solaris
• Windows XP/2000/NT

• Why is it called 'Andrew' File System?
Andrew was the name of a joint project between CMU and IBM to set up a distributed computing environment on the CMU campus. It was named after the founders of the university, Andrew Carnegie and Andrew Mellon. Later, as the file system became common over the country, the engineers tried to change the name. At that time, however, changing the root of the file system, /afs, had to involve a lot of work for many AFS sites who had set everything up under /afs. Thus, AFS remained as it is now.