NAMD / pypelinin

Python library to distribute jobs and pipelines among a cluster
3 stars 5 forks source link

Broker should only ask for jobs it can execute #46

Open turicas opened 11 years ago

turicas commented 11 years ago

Currently, Router announces that there is a new job, but it does not say which worker Broker should execute for this job until Broker asks for the job. If Broker doesn't have the worker for it, the job will fail. So, I propose:

Optionally, Broker can send all workers it has loaded to Router when it starts, so Router will know all workers it can execute in the entire cluster.

turicas commented 11 years ago

This issue is a bit related to #29.