June 1998课件.pptVIP

Optimized Execution of Modules M1 M4 (M1+M2) M5 M2 M3 (M1+M2) i1 e1 e4 e3 e2 i3 i4 i5 i2 e5 time M1 M4 M5 M2 M3 i1 e1 e2 e3 e4 e5 i2 i3 i4 i5 time data dependencies execution of a module non-optimized optimized by scheduler according to estimates invoke a method i e extract results 精品文档 Decomposed Parallel Execution time M1 M4 (M1+M2) M5 M2 M3 M1+M2) optimized by scheduler according to estimates invoke a method extract results set up / set attributes Long setup times occur, for instance, when a subset of a large database has to be loaded for a simple search, say Transatlantic fights for an optimal arrival. 精品文档 M1 M4 (M1+M2) M5 M2 M3 (M1+M2) Decomposed Optimized Execution M1 M4 (M1+M2) M5 M2 M3 (M1+M2) optimized by scheduler according to estimates invoke a method extract results set up / set attributes time prior time 精品文档 M1 M4 (M1+M2) M5 M2 M3 (M1+M2) Repeated invocations M1 M4 (M1+M2) M5 M2 M3 (M1+M2) optimized by scheduler according to estimates invoke a method extract results set up / set attributes time prior time 精品文档 M1 M4 (M1+M2) M5 M2 M3 (M1+M2) Repeated Extractions M1 M4 (M1+M2) M5 M2 M3 (M1+M2) optimized by scheduler according to estimates invoke a method extract results set up / set attributes time prior time 精品文档 Scheduling: Simple Example 1 cost_ground_ih = cost_mmh.INVOKE (Cost_for_Ground, 1 List_of_City_Pairs = city_pairs,Goods = info_goods) 2 WHILE (cost_ground_ih.EXAMINE() != DONE) {} 3 cost_list_ground = cost_ground_ih.EXTRACT() 3 cost_air_ih = cost_mmh.INVOKE (Cost_for_Air, 2 List_of_City_Pairs = city_pairs,Goods = info_good) 4 WHILE (cost_air_ih.EXAMINE() != DONE) {} 4 cost_list_air = cost_air_ih.EXTRACT() order in unscheduled megaprogram order in automatically prescheduled megaprogram 精品文档 Scheduling: Possible Actions INVOKES: call INVOKE’s as soon as possible may depend on other data moving it outside of an if-block: depending on cost-function (ESTIMATE of this and followi