The cost of reconciling consistency and state manage- ment with high availability is highly magnified by the un- precedented scale and robustness requirements of today’s Internet applications. We propose two strategies for im- proving overall availability using simple mechanisms that scale over large applications whose output behavior toler- ates graceful degradation. We characterize this degradation in terms of harvest and yield, and map it directly onto engi- neering mechanisms that enhance availability by improving fault isolation, and in some cases also simplify program- ming. By collecting examples of related techniques in the literature and illustrating the surprising range of applica- tions that can benefit from these approaches, we hope to motivate a broader research program in this area.