We need to run the failure detector in the protocol, hence we need to run the introducer from MP2 in the background.
- On VM1, run
python introducer.py &> introducer.log &. The introducer listens for machines to connect to it.
- On VM1 and VM2, run
python nimbus.pyto run the master and backup master
- On VM3, run the crane client (used to submit jobs) as
python crane-client.py. - At the prompt, use start config_file_name to submit a Crane job
- On the remaining VMs, run
python worker.py. The workers then listen for jobs from the Nimbus