Download presentation
Presentation is loading. Please wait.
Published byJoel Powell Modified over 8 years ago
1
Lu Qin Center of Quantum Computation and Intelligent Systems, University of Technology, Australia Jeffery Xu Yu The Chinese University of Hong Kong, China Lijun Chang The University of New South Wales, Australia Hong Cheng The Chinese University of Hong Kong, China Xuemin Lin The University of New South Wales, Australia East China Normal University, China Mahmoud Agbareya, 13 January 2015
2
Agenda Introduction The Scalable Graph Processing Class ( SGC ) SGC Algorithms Performance Studies 2
3
Agenda Introduction The Scalable Graph Processing Class ( SGC ) SGC Algorithms Performance Studies 3
4
Introduction What is MapRecuce? MapReduce Class ( MRC ) Minimal MapReduce Class ( MMC ) 4
5
Introduction What is MapRecuce? MapReduce Class ( MRC ) Minimal MapReduce Class ( MMC ) 5
6
What is MapReduce Programming model for processing large data sets in distributed systems. Process the data as (key, value) pairs. May executes in rounds Each round has three phases: map, shuffle and reduce. Each round runs in many machines – each machine is dedicated for one task (map or reduce) Introduced by two researchers from Google in 2004. Most popular implementation is Hadoop. 6
7
What is MapReduce (cont.) Example 7
8
Introduction What is MapRecuce? MapReduce Class ( MRC ) Minimal MapReduce Class ( MMC ) 8
9
MapReduce Class ( MRC ) Definition 9
10
10 MapReduce Class ( MRC ) (Graph version) Definition
11
Introduction What is MapRecuce? MapReduce Class ( MRC ) Minimal MapReduce Class ( MMC ) 11
12
Minimal MapReduce Class ( MMC ) Definition 12
13
13 Minimal MapReduce Class ( MMC ) Definition
14
Agenda Introduction The Scalable Graph Processing Class ( SGC ) SGC Algorithms Performance Studies 14
15
The Scalable Graph Processing Class ( SGC ) Motivation Preliminaries SGC Definition Two graph operators in SGC : NE Join EN Join 15
16
The Scalable Graph Processing Class ( SGC ) Motivation Preliminaries SGC Definition Two graph operators in SGC : NE Join EN Join 16
17
Motivation 17
18
Motivation We aim to define a MapReduce class in which, graph algorithm has the following three properties: Scalability: The algorithm can always be speeded up by adding more machines. Stability: The algorithms stops in bounded number of rounds. Robustness: The algorithm never fails regardless of how much memory each machine has. 18
19
The Scalable Graph Processing Class ( SGC ) Motivation Preliminaries SGC Definition Two graph operators in SGC : NE Join EN Join 19
20
Preliminaries 20
21
Preliminaries 21
22
Preliminaries 22
23
The Scalable Graph Processing Class ( SGC ) Motivation Preliminaries SGC Definition Two graph operators in SGC : NE Join EN Join 23
24
24 Scalable Graph Processing Class ( SGC ) definition
25
25 Scalable Graph Processing Class ( SGC ) definition
26
The Scalable Graph Processing Class ( SGC ) Motivation Preliminaries SGC Definition Two graph operators in SGC : NE Join EN Join 26
27
Two graph operators in SGC 27
28
The Scalable Graph Processing Class ( SGC ) Motivation Preliminaries SGC Definition Two graph operators in SGC : NE Join EN Join 28
29
NE Join 29
30
NE Join 30
31
NE Join in MapReduce 31
32
NE Join in MapReduce 32
33
The Scalable Graph Processing Class ( SGC ) Motivation Preliminaries SGC Definition Two graph operators in SGC : NE Join EN Join 33
34
EN Join 34
35
EN Join 35
36
EN Join 36
37
EN Join in MapReduce 37
38
Agenda Introduction The Scalable Graph Processing Class ( SGC ) SGC Algorithms Performance Studies 38
39
SGC Algorithms Basic Graph Algorithms: Breadth First Search Page Rank Graph Keyword Search Advanced Algorithms: Connected Component Minimum Spanning Forest 39
40
SGC Algorithms Basic Graph Algorithms: Breadth First Search Page Rank Graph Keyword Search Advanced Algorithms: Connected Component Minimum Spanning Forest 40
41
Breadth First Search 41
42
SGC Algorithms Basic Graph Algorithms: Breadth First Search Page Rank Graph Keyword Search Advanced Algorithms: Connected Component Minimum Spanning Forest 42
43
Page Rank 43
44
Page Rank 44
45
SGC Algorithms Basic Graph Algorithms: Breadth First Search Page Rank Graph Keyword Search Advanced Algorithms: Connected Component Minimum Spanning Forest 45
46
Graph Keyword Search 46
47
Graph Keyword Search 47
48
SGC Algorithms Basic Graph Algorithms: Breadth First Search Page Rank Graph Keyword Search Advanced Algorithms: Connected Component Minimum Spanning Forest 48
49
49 Forest Initializing Conditional Star Hooking Unconditional Star Hooking Pointer Jumping Star Detection Procedure
50
Connected Component 50 Forest Initializing: Line 1: find the minimum neighbor for each node and set it to be the parent.
51
Connected Component 51 Forest Initializing:
52
Connected Component 52
53
Connected Component 53 Forest Initializing:
54
Connected Component 54 Star Detection: Rules to detect that node is not in star (applied in order)
55
Connected Component 55
56
Connected Component 56
57
Connected Component 57
58
Connected Component 58
59
Connected Component 59 Conditional Star Hooking (inside the loop):
60
Connected Component 60 Conditional Star Hooking (inside the loop): After Conditional Star Hooking, it’s guaranteed that there are no edges between two starts.
61
Connected Component 61
62
Connected Component 62 Unconditional Star Hooking (inside the loop):
63
Connected Component 63
64
Connected Component 64 Pointer Jumping (inside the loop):
65
Connected Component 65
66
SGC Algorithms Basic Graph Algorithms: Breadth First Search Page Rank Graph Keyword Search Advanced Algorithms: Connected Component Minimum Spanning Forest 66
67
Minimum Spanning Forest 67
68
Minimum Spanning Forest 68
69
69 Forest Initializing Cycle Breaking Edge Hooking Pointer Jumping
70
Minimum Spanning Forest 70
71
Minimum Spanning Forest Forest Initialization 71
72
Minimum Spanning Forest 72
73
Minimum Spanning Forest Cycle Breaking 73
74
Minimum Spanning Forest Pointer Jumping 74
75
Minimum Spanning Forest 75
76
Minimum Spanning Forest 76
77
Minimum Spanning Forest 77
78
Minimum Spanning Forest 78
79
Minimum Spanning Forest Edge Hooking 79
80
Minimum Spanning Forest 80
81
Agenda Introduction The Scalable Graph Processing Class ( SGC ) SGC Algorithms Performance Studies 81
82
82
83
83
84
84
85
85
86
86
87
87
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.