Description
In this project, you are going to enhance your CORBA implementation of the Distributed
Player Status System (DPSS) developed in Assignment #2 to be software failure tolerant
or highly available using process replication.
This project is suitable for teams of 4
students, as described in the following:
For this project, extend your Distributed Player Status System (DPSS) implementation
from Assignment #2 to tolerate a single software (non-malicious Byzantine) failure using
active replication.
Your actively replicated DPSS server system should have at least three replicas each
running a different implementation (in different hosts on the network).
One of the replicas in the group is the designated leader and receives requests from
clients through a CORBA Front End (FE). The leader of the server group broadcasts a
client request (received through the front end) atomically to all the other replicas in the
group using a reliable FIFO broadcast mechanism, receives the responses from them and
sends a single correct response back to the client (through the front end).
The leader also
informs the Replica Manager (RM) if a replica produces incorrect result so that the RM
can replace a failed replica that produced incorrect result three times successively with
another good one.
Since the entire server system (Replicas, FE, and RM) usually runs on a local area
network, the server replicas, FE, and RM communicate among them using the unreliable
UDP protocol. However, this communication should be made reliable and FIFO in order
to avoid message loss and guarantee correctness.
Specifically do the following:
Assuming that server failure could be due to software bugs (i.e. non-malicious
Byzantine failures), design your actively replicated, software failure-tolerant
server system;
Modify the server implementation from Assignment #2 so that it can work either
as the group leader or as a regular server. The group leader receives a request
from the CORBA front end, FIFO broadcasts the request to all the server replicas,
receives the responses from the server replicas and sends a single correct response
back to the client. The regular server receives requests from the leader, executes
the request and sends the response back to the leader;
Design and implement the front end (FE) which receives a client request as a
CORBA invocation, forwards the request to the leader process, receives the result
from the leader and sends it back to the client;
Design and implement the Replica Manager (RM), which creates and initializes
the actively replicated server subsystem. The RM also manages the server replica
group information, maintains if a replica has produced incorrect result and
replaces a failed replica that produces incorrect result three times successively
with another good one;
Design and implement a reliable FIFO broadcast subsystem over the unreliable
UDP layer;
Integrate all the modules properly, deploy your application on a local area
network, and test the correct operation of your application using properly
designed test runs. You may simulate a software failure by returning a random
result.
Submitting Project
– Naming convention for zip file: Create one zip file, containing all source files
(.java, .doc or .pdf or .txt, etc.) for your project using the following naming
convention:
The zip file should be called P_studentID, where # is the number
of the project studentID is your student ID number. For example,
for the project, student 123456 would submit a zip file named
P_123456.zip.
– Submit your zip file at: https://fis.encs.concordia.ca/eas/ as Project and
submission. The project submitted to the wrong directory would be discarded and
no replacement submission will be allowed.
– Submit only ONE version of a project. If more than one version is submitted the
last one, before the deadline date, will be graded and all others will be
disregarded.
Evaluation Criteria of the project (100 points)
Activities Points
Design Documentation:
– Describe and explain your design and architecture clearly, including
theories (protocol, algorithm) you apply, how to implement. [20 pts.]
– Design proper and sufficient test scenarios, which should include
testing data and results. [10 pts.]
30 pts.
Q2: Demo on Zoom platform:
– Introduce your application architecture: [20 pts.]
– Demo your designed testing scenarios to illustrate the correctness of
your design. If your testing scenarios do not cover all possible issues,
you’ll lose part of mark up to 40%. [50 pts.]
70 pts.
