DSS_py

DSS_py is a Distributed System Simulator. This library allows one to simulate a distributed environment in a single system. One can create a number of machine instances, assign them tasks in the form of functions and make them communicate between each other through messages.

Get Started

1. Download the files and copy dss.py, imports.py and config.ini to the folder where you want to work.
2. Set the maximum number of machine instances you want to work with in config.ini.
3. Create a file named functions.py and keep the functions that you want to assign as tasks to the machines.
4. Create another file and start coding. Do include the line from dss import * in your main code (this file).

Go through the demos in the Demo folder for implemented examples.

Documentation

Creating a New Instance

machine_1 = machine()

Assigning a Task in the Form of a Function

machine_1.execute_func("func_name", 1, 2)

Where func_name is the name of the function, followed by the arguments that the function accepts.

Every function that is to be assigned as a task to the machine instance has to have their first parameter as the identity variable. One can name it whatever he/she wants. This identity variable has to be used while calling other functions such as send(), recv() and get_machine_id(). The function can have its formal variables (as sent through execute_func()) just after the identity variable.

For example, lets say I want to create a machine instance m1 and assign it a task of printing 10 numbers.

Keep the function definition in functions.py

def foo(identity_variable, x):
    for i in range(x):
        print i
# just a note, this identity_variable has to be used to call send(), recv() and other further defined functions

Creating machine instance and assign task through function

>> from dss import *
>> m1 = machine()
>> m1.execute_func("foo", 10)

NOTE: The machine instance would run the function in the background. One can create another machine instance just after the above lines and assign some function.

Sending a Message to another Machine Instance

identity_variable.send("machine_1", "hello", block)

Where "machine_1" is the machine ID of the destination machine and "hello" is the message. send() can be made blocking/non-blocking by using the block variable. Setting block to 0, would make the corresponding send() non-blocking, all other values would make it blocking.

send() returns 1 on success.

send() returns 0 when block is set to 0 (send() is marked non-blocking) and the requested operation would block.

send() returns -1 when unsuccessful. This error will occur when you pass machine id that is not present.

Receiving a Message from other Machine Instance

identity_variable.recv(block)

recv() will get any message that has been sent to the corresponding machine. recv() can be made blocking/non-blocking by using the block variable. Setting block to 0, would make the corresponding recv() non-blocking, all other values would make it blocking.

recv() returns message, sender_machine_id on success.

recv() returns 0, 0 when block is set to 0 (recv() is marked non-blocking) and the requested operation would block.

recv() returns -1, -1 when unsuccessful.

Get the current machine ID

identity_variable.get_machine_id()

Returns the machine id of the present machine. It returns in the form of "machine_1".

Contact

Feel free to mail me with any queries at [email protected]