Model Components

Basic building components (generators/processes) for parallel models. In general, the framework allows one to build a model by defining an arrival, routing, and job/servicing process such that work is introduced in the order of Arrivals->Router->Job/Servicing.

class parallelqueue.network.Network(**kwargs)

The Network constructor allows a user to create a queueing network by overriding each member. Upon generation, jobs flow through a network in the order of: Arrivals → Router → Job. By default, the Network class can be used to handle JSQd, Redundancy-d, and Threshold-(d,r) models with general arrival and service distributions.

Arrivals(system, env, number, queues, **kwargs)

This generator/process defines how jobs enter the network

static Job(system, env, name, arrive, queues, choice, **kwargs)

This generator/process defines the behaviour of a job (replica or original) after routing.

Router(system, env, name, queues, **kwargs)

This generator/process specifies the scheduling system used.

parallelqueue.arrivals.DefaultArrivals(router, system, env, number, queues, **kwargs)

General arrival process; interarrival times are defined by the given distribution

Parameters

  • router – Router process.

  • system (base_models.ParallelQueueSystem) – System providing environment.

  • env (simpy.Environment) – Environment for the simulation

  • number (int) – Max numberJobs of jobs if infiniteJobs is false.

  • queues (List[simpy.Resource]) – A list of all queues making up the parallel system.

parallelqueue.routers.DefaultRouter(job, system, env, name, queues, **kwargs)

Specifies the scheduling system used. If replication is enabled, this router tracks each set of replicas using a base_models.ParallelQueueSystem.ReplicaDict which can be accessed by network.Network.Job processes.

Parameters

  • job – Job process.

  • system (base_models.ParallelQueueSystem) – System providing environment.

  • env (simpy.Environment) – Environment for the simulation.

  • name (str) – Identifier for the job.

  • queues (List[simpy.Resource]) – A list of queues to consider.

parallelqueue.routers.NoInSystem(R)

Total number of Jobs in the resource R.

parallelqueue.routers.QueueSelector(d, parallelism, counters)

The actual queue selection logic.

parallelqueue.jobs.DefaultJob(system, env, name, arrive, queues, choice, **kwargs)

For a redundancy model, this generator/process defines the behaviour of a job (replica or original) after routing.

Parameters

  • system (base_models.ParallelQueueSystem) – System providing environment.

  • env (simpy.Environment) – Environment for the simulation

  • name (str) – Identifier for the job.

  • queues (List[simpy.Resource]) – A list of queuesOverTime.

  • arrive (float) – Time of job arrival (before replication).

  • choice (int) – The queue which this replica is currently in