Lab3

Local cluster

1. Dependencies

   Before running the code, you need to ensure that you have the necessary dependencies installed. These include:

   • Python
   • Ray
   • Make

2. Start Ray cluster

   You need to start a Ray cluster to run the code. To start the Ray cluster in head mode, run the following command inside local-cluster directory:

   make run-head
   

   This will start the Ray head node with the Ray dashboard enabled.

3. Start Ray workers

   After starting the Ray cluster in head mode, you need to start Ray workers. Run the following command inside local-cluster directory to start Ray workers:

   make run-worker ADDRESS=<head-node-ip-address>
   

   Replace with the IP address of the machine running the Ray head node. You can start as many workers as needed.

4. Run tasks

   There are several tasks that can be run, including tasks involving actors and objects. To run a task, use one of the following commands inside local-cluster directory:

   make run-task
   make run-objects
   make run-actors1
   make run-actors2
   make run-actors3
   

Dockerized cluster

1. Dependencies

   Before running the code, you need to ensure that you have the necessary dependencies installed. These include:

   • Python
   • Docker

2. Run docker container

   Run the following command inside docker-cluster directory

   docker-compose up
   

3. Run tasks

   There are several tasks that can be run, including tasks involving actors and objects. To run a task, use one of the following commands inside docker-cluster directory:

   python ray-lab-1-task.py
   python ray-lab-2-objects.py
   python ray-lab-3-actors-1.py
   python ray-lab-3-actors-2.py
   python ray-lab-3-actors-3.py