This experiment simulates a distributed system with a set of nodes that need to agree on a common value. Some of these nodes can be Byzantine, meaning they can behave maliciously.

Simulation Controls

The simulation provides the following controls:

• Number of Nodes (n): You can set the total number of nodes in the system. The system requires n > 3t to be fault-tolerant, where t is the number of Byzantine nodes.
• Number of Byzantine Nodes (t): You can set the number of nodes that will exhibit Byzantine behavior. These nodes will send conflicting or incorrect messages to other nodes.
• Client Request: You can initiate a request from a client to the primary node. The request will contain a value that the nodes need to agree upon.
• Run Simulation: This button starts the simulation of the PBFT algorithm.

Steps to run the experiment

1. Set the number of nodes (n) and Byzantine nodes (t). Ensure that n > 3t. For example, you can start with n=4 and t=1.
2. Initiate a client request. The client will send a request to the primary node (Node 0).
3. Run the simulation. Observe the messages being passed between the nodes in the simulation log.
   • The primary node sends a pre-prepare message to the backup nodes.
   • The backup nodes send prepare messages to each other.
   • Once a node receives 2t matching prepare messages, it sends a commit message.
   • Once a node receives 2t+1 commit messages, it executes the request and sends a reply to the client.
4. Observe the behavior of Byzantine nodes. The Byzantine nodes will be highlighted, and their malicious messages will be visible in the log. For example, a Byzantine node might send a "retreat" message when the correct message is "attack".
5. Verify the outcome. Despite the presence of Byzantine nodes, the loyal nodes should be able to reach a consensus on the correct value and send the correct reply to the client. The client will wait for t+1 identical replies before accepting the result.
6. Experiment with different values of n and t. Try to set n <= 3t (e.g., n=3, t=1). Observe that the system is unable to reach a consensus in this case. This demonstrates the n > 3t condition for Byzantine fault tolerance.

What to Observe

• The sequence of messages (pre-prepare, prepare, commit, reply).
• How loyal nodes handle messages from Byzantine nodes.
• The final consensus reached by the loyal nodes.
• The conditions under which the system fails to reach a consensus.