Defining an alerting strategy

Asking the question What is our alerting strategy? will likely result in a bunch of blank stares. We spend so much time on creating monitoring dashboards and then often fail to take the next step and activate the data through alerts. Or, we go the other direction and alert on so many things that the alerts become meaningless noise. Defining an alerting strategy is about taking a planned and governed approach to determine what the system should monitor, by what method, how frequently, and what actions are expected. It is suggested that the strategy is created outside of the tool first, and then implemented in the toolset. This discipline forces strategic thinking and allows for rationalization of the number of conditions to be monitored. Keep in mind that a strategy, by design, is a high-level blueprint and will not be specific enough to cover all implementation details. The specific elements involved in defining a good strategy are contained in the Alert...