How to forget a customer, if all resources are being used at the time

Question

I have a Server that can serve at-most 3 customers at a time. I have modeled it as a SimPy Resource with capacity=3 as follows. How can I update my code such that whenever a new customer arrives, if all of the server's capacity is being utilized, the new customer is simply forgotten?

I have read the documentation, but I cannot find an example that fits this scenario.


SERVERS= 3

class Servers:

    def __init__(self, env):
        self.cashier = simpy.Resource(env, capacity= SERVERS)

def Customer(name, env, servers):
    print('Customer %s arriving at %s' % (name, env.now))
    with servers.cashier.request() as req:
        yield req
        print('Customer %s starts being served at %s' % (name, env.now))
        yield env.timeout(35)
        print('Customer %s done served at %s' % (name, env.now))

def customer_generator(env, servers):
    for i in range(100):
        env.process(Customer(i, env, servers))
        yield env.timeout(5)

env = simpy.Environment()
servers = Servers(env)
cus_generator = env.process(customer_generator(env, servers))
env.run(1000)

score 0 · Answer 1 · answered Aug 13 '19 at 12:52

SimPy offers a simple way to check how much of a resource's capacity is currently being used. This can be done using <resource>.count (link to API).

Your example can be modified as follows:

SERVERS= 3

class Servers:

    def __init__(self, env):
        self.cashier = simpy.Resource(env, capacity= SERVERS)

def Customer(name, env, servers):
    print('Customer %s arriving at %s' % (name, env.now))

    # check if any of the cashiers are free:
    if servers.cashier.count < servers.cashier.capacity:
        with servers.cashier.request() as req:
            yield req
            print('Customer %s starts being served at %s' % (name, env.now))
            yield env.timeout(35)
            print('Customer %s done served at %s' % (name, env.now))
    else:
        print('Customer %s leaving as there was no cashier free to serve them' % (name))

def customer_generator(env, servers):
    for i in range(100):
        env.process(Customer(i, env, servers))
        yield env.timeout(5)

env = simpy.Environment()
servers = Servers(env)
cus_generator = env.process(customer_generator(env, servers))
env.run(1000)

Note that the customer process gets destroyed (forgotten) automatically in the else branch as the function ends without hitting another yield.

How to forget a customer, if all resources are being used at the time

1 Answers1