Detecting failures in distributed systems with the Falcon spy network
Joshua B. Leners, Hao Wu, Wei-Lun Hung, Marcos K. Aguilera, Michael Walfish · 2011
A common way for a distributed system to tolerate crashes is to explicitly detect them and then recover from them. Interestingly, detection can take much longer than recovery, as a result of many advances in recovery techniques, making failure detection the dominant factor in these systems' unavailability when a crash occurs.