The updating of node's data is currently goberned by the Manager that trigger periodically, however this leads to results to be held in the DB untill they can be processed.
The manager should be divided into two workflows: one that triggers new tasks periodically and one that processes results and updates nodes.
The later workflow can be then directly triggered by the Evaluator as soon as the result is written.
The updating of node's data is currently goberned by the Manager that trigger periodically, however this leads to results to be held in the DB untill they can be processed. The manager should be divided into two workflows: one that triggers new tasks periodically and one that processes results and updates nodes. The later workflow can be then directly triggered by the Evaluator as soon as the result is written.