Random initialization of the logical clock of individual processes and usage of Berkeley’s algorithm to synchronize these clocks to the average clock. Implementation of the causal ordered multicasting for the distributed system.
Part I: Berkeley Algorithm
File named Berkeley.cpp has the code for Part I. It can be compiled as below:
g++ -std=c++11 -o berkeley Berkeley.cpp -pthread
./berkeley process 1
./berkeley process 2
./berkeley process 3
./berkeley master
All client processes are executed before runnin the master program. Those client processes would be waiting for the master server to connect. Once it starts, all waiting processes would start running and generate a random logical clock for themselves. Those clock values are used further for clock synchronization.
Part II: Causal Ordering
File named CausalOrdering.cpp has the code for Part II. The non-causal ordering part is implemented in the file named NonCausalOrdering.cpp. These files can be compiled as below:
g++ -std=c++11 -o causalordering CausalOrdering.cpp -pthread
./causalordering 0
./causalordering 1
./causalordering 2
g++ -std=c++11 -o noncausalordering NonCausalOrdering.cpp -pthread
./noncausalordering 0
./noncausalordering 1
./noncausalordering 2
Process 0 is the first process, process 1 is the second process and process 2 is the third process. All these processes send messages to each other and themselves as well. CausalOrdering.cpp program ensures that all the messages are reached in a causal ordered manner whereas NonCausalOrdering.cpp program displays how the messages are delivered without the implementation of causal ordering.
Part III: Distributed Lock
This part is combined with the Part I itself.
Locks are used for the shared file and the file is read by all the processes and the counter value present in the file is updated accordingly.