Method for providing database recovery across multiple nodes

摘要: A three-phase database crash recovery mechanism is detailed including a forward phase, backward and third, so-called “sideward” phase for of transactions which were interrupted at the time crash, using Global Transaction IDs to track status coordinating participating nodes. Depending upon transaction either query message node or vote nodes generated processed, thereby allowing most be completed. Additional implementations are provided recovery, without duplication efforts, across multiple in parallel environment, cascaded wherein local triggered by remote system, concurrent started concurrently both coordinator participant