As an ONAP Policy Developer, I would like to ensure that I can choose how any pair of control loops that target a common network element will interact. E.g.,

• "Both control loops A & B are allowed to operate on a common network element simultaneously."
• "If control loop A is operating on a network element, control loop B is prohibited for making changes until A has finished and 5 minutes have elapsed."
• "If control loop A is operating on a network element, and control loop B requests to operate on that element, control loop A's operations are reverted, after which B is permitted to operate on the network element."

To that end, this epic aims to ensure:

• I can easily specify coordination behavior using a straightforward and high-level API.
• I can insert new coordination directives during runtime.
• I can coordinate between closed loops in which a single event received by the policy segment of either control loop may target operations on several network elements.
• My coordination directives operate correctly across large geographic regions, many control loops, or multiple controller instances.
• Uncommon race conditions are correctly handled: e.g., garbage collection of coordination state when a control loop fails.