COS 318 : Operating system Fall 2002, Princeton University Project 3: Filesystem

In Project 1, you grabbed data from the standard Unix filesystem when generating pages for your web server. Now, you get to build the backing filesystem yourself and (slightly) modify your server to interact with it. This project is not conceptually difficult, but will probably be a fair amount of coding.

High-Level Details

• Rather than write to a real disk, you'll be building a disk inside of a large file. We'll provide the functions necessary to create this "disk", load it from disk, and read/write sectors. This approach allows you to do all of the development in user space and use whatever debugging facilities that you've been using already
• We'll provide a new client that introduces some new methods. Previously, your server had to handle only the GET method for retrieving files. This new client will be capable of issuing (our custom) methodsfor creating directories, deleting files/directories, and storing files or parts of files. What this means is that you'll be able to use our client to create a populated directory tree inside of your disk, and then use the web server to browse this content. You'll have to understand at least one new header with one of the new methods.
• You'll have to write versions of various system calls that you used in Project 1 - these will include versions of open, close, read, and write. Additionally, you'll also have to implement versions of stat, mkdir, and a few others. You won't have to handle all of the various flags and parameters, but should implement enough to get the project working correctly. For example, the open() system call can be used to try to test the existence of a file in addition to creating it for writing. You will need to implement at least the file creation flag in order to properly handle our "put" method. However, you can't always blindly assume you'll want to create, since you may receive a GET for a non-existent file.
• Finally, you'll also have to implement enough of the calls related to getting file metadata and directory information in order to properly build directory listings

Submit your final solution electronically using the following command:

/u/cs318/bin/318submit 3 README Makefile (list of files)

The submit command copies your files to the directory /u/cs318/submit/UserLogin/number and lists all the files that you have submitted for assignment number. UserLogin is your user account name. If you execute submit after the project deadline, your files are placed in directory number_late. You can run submit more than once, and you can submit partial lists of files.

Each group needs to submit only one copy.  This means that only one of you needs to submit.

Hints

Within the filesystem, start with the creation of files and try to get that working properly. Then build directories on  top of that mechanism, and finally work on directory listing support.  Not only is this a logical way of approaching the project, but in the event that something doesn't work, hopefully you'll get as much partial credit as possible.

Secondly, don't make the mistake of trying to change the high level interfaces too much. Assume you have a working server or can get a good reference server from Project 1, and then just introduce wrappers for the relevant system calls. By having wrappers that can either call the real call or your replacement, you'll be able to do testing quickly and efficiently. If you change the high-level interfaces, you won't have this luxury.

Finally, if you're feeling brave and want to allow your partner access to files, you should be able to play around with the unreadable directory tricks we discussed in class. You can create a strangely-named intermediate directory which leads to the real, readable directory. You can make the files in this directory world-writeable, so unless someone else got hold of your strange intermediate directory name, it's unlikely that anyone else will be able to modify them.

CS318 Staff