Design for handling events/alerts
We need to come up with a way to have a generic way of handling events with potential reactions to them.
Although it's not immediately urgent (can be done after 0.8 if need be), the sooner we have an idea how the event handling will look like in the future, the better.
At the IETF, there is a concept called ECA (Event-Condition-Action). The idea is that you can define what event you are looking for (e.g. server down or duplicate address detected), what conditions have to be met (always, or if repeated more than 3 times in last hour), then some action defined is executed (restart server or send e-mail warning about duplicates).
Some references worth looking at:
NOTE: Many of those are associated with YANG/NETCONF. We do not want to use those technologies in Stork. However, we could potentially reuse the concepts.