Quote: optimistic recovery for distributed system via repeatable, message-driven recovery units; reconstruct consistent state after failure

Quotation Skeleton

In this paper we describe the optimistic recovery … a distributed system. … [p. 206] Instead of consistent checkpoints … , enough information is saved to reconstruct a … [p. 207] The logical machine is partitioned into a … By allowing each recovery unit to schedule multiple … We make the following assumptions … Reliable, FIFO channels between recovery units … Stable storage … Spare processing capacity. It is always possible to … previously logged recovery information on stable storage. … We assume the behavior of each recovery unit …   Google-1   Google-2

Copyright clearance needed for quotation.

Related Topics

Topic: reliability of distributed systems (33 items)
Topic: optimistic update for concurrency control (35 items)