In modern business, the cloud computing processes that manage business logic—for example, the sequence of bank transactions and shipment logistics necessary to fulfill an online order—are so intertwined that a single failure can cascade and stop business in an entire geographic region. If not resolved within seconds or minutes, a system "downtime" event can cost an enterprise millions of dollars.<br><br>Downtime is a program comprising eight "worker" samples that initially keep time with a metronomic grid. Each worker has the power to digress from its schedule and pull the others with it, their aggregate disobedience skewing the grid. The chart interface that watches the system crashing, recovering, and crashing again is inspired by observability dashboards used to monitor the reliability and performance of business systems.<br><br>The Feral File versions single out individual workers. Each performs a single routine that charts a position somewhere between being available and unavailable for work.<br><br>"Downtime" also connotes rest, or relief from the chorus that demands worker availability, access, and attention.