Category: Distribution System Project Ideas

Distributed system on One Lane Bridge project idea is to provide solution for vehicles coming from different  direction to cross one lane bridge.

Develop a server process to manage use of the bridge. Assume the cars are client processes that can remotely get connect to the server at any time.

In your simulation program, have new cars join in the session remotely and each car repeatedly try to cross the bridge. They should spend a random amount of time on the bridge and should delay for a random amount of time before crossing again. Each car client can specify some parameters such as its speed, average delaying time after crossing, initial heading direction etc. Ideally, each client should get a graphical interface showing the current status on the bridge utilization.

Implement a stock market simulation game for several human players and several computer players.You are to create a fast-paced on-line interactive game of stock market, which has many of the characteristics of electronic commerce. The entities in the game are a bank, a stock exchange, several companies and several players. The time between events has been significantly accelerated to make the game more challenging and interesting.

Bank: The one bank holds the cash of any player.
A player can request a new account, deposit and withdraw money from the account. The bank has a maximum of 10 accounts. Once every minute during the game, all money in an account accumulates 1% interest.

Company: A company has a name and issues stock. Once a minute, the company’s price of the stock is adjusted based on a random process and the law of supply and demand. Initially each company has 1000 shares of stock, which starts at $30 a share. There are several companies for example three to ten in the game. A company buys or sells shares to a player by accepting or rejecting the player’s bid. The process of deciding to accept or reject may be random or based on sound business practices.

Player: There can be two types of players – human and computer. A player starts with $1000 and buys and sells stock from companies or other players by trading through the stock exchange. The goal of a player is to make money. A player buys or sells shares from another player or a company by posting bids on the stock exchange. The company or the other player may accept or reject the player’s bid. The process of deciding how the computer-based player accepts or rejects may be random or based on sound business practices. There are two to ten players.

Stock Exchange: Players and companies must register with the stock exchange. All trading of stocks must go through the centralized stock exchange. The stock exchange posts current bids and the prices of recent transactions on each stock.

The winner is the player with the highest total value in cash and stock value after a certain period of time. Of course, a player can not buy shares of stock that don’t exist and he or she must have enough cash to purchase the desired shares. The bank, stock exchange, companies and computer players are trusted entities, i.e., they don’t cheat.

Replicated file Distributed system project idea explains a simple way to increase the likelihood that a critical data file is always accessible is to keep a back-up copy of the file on another disk, usually one that is attached to a different machine. The file system provides transparent replication. In particular, suppose there are n copies of a data file and n server modules. Each server provides access to one copy of the file. A client interacts with any one of the server modules. The servers interact with each other to present clients with the illusion that there is a single copy of the file. Each file exports four operations: open, close, read and write for client. The file servers interact with each other to ensure that copies of the file are kept consistent and that at most one client at a time is permitted to write into the file. Each file servers has a local lock manager process that implements a solution to the readers/writer problem. Some points to bear in mind.

a. Each fileserver exports two sets of operations: those called by its clients and those called by other file servers. Each server module keeps track of current access mode by using a lock manager. For example,the file is not written if it was opened for reading. Multiple readings however are allowed. If a certain operation needs to get permission from all lock managers, please acquire the locks in the same order for all clients. Otherwise, deadlock may occur.

b. Within write procedure, a module first update its local copy then concurrently updates every remote
copy. It is analogous to using a write-through cache update policy. An alternative approach would be to
update the remote copy when the file is closed.

Demo purpose: Store duplicated simple text files in several fileservers for demo purpose. Write operation
just simply appends more text to the file. Client gets connected to any fileserver in a graphical user interface.