Andrew Yongjoon Kong, profile picture
Uploaded byAndrew Yongjoon Kong
PPTX, PDF222 views

Graph analysis platform comparison, pregel/goldenorb/giraph

The document discusses a system for large-scale graph processing, focusing on Google's Pregel architecture, alongside its alternatives like GoldenOrb and Giraph. It highlights the limitations of MapReduce for iterative graph processing and proposes Pregel as a more efficient model that maintains graphs in memory and reduces network communication. The document also outlines the framework's fault tolerance and architecture, emphasizing the master/worker model for computation management.

Embed presentation

Download to read offline
A System for Large-Scale Graph Processing Pregel, GoldenOrb, Giraph 2012-07-18 Andrew Yongjoon Kong sstrato.open@gmail.com 1
Contents • Introduction • Model of Computation • Pregel Architecture • Goldenorb • Implementation • Future work 2
Introduction 3
Introduction • Today, Many practical computing problems concern large graphs • Applied algorithms - Shortest paths computations - Page rank - Clustering techniques • MapReduce is ill-suited for graph processing - Many iterations are needed for parallel graph processing - Materializations of intermediate results at every MapReduce iteration harm performance 4
Introduction • Hadoop is well-suited for non-iterative, data parallelized processing 5 Smith Waterman is a non iterativ e case and of c ourse runs fine
Introduction 6 map map reduce Compute the dist ance to each dat a point from eac h cluster center a nd assign points to cluster centers Compute new cluster centers Compute new clust er centers User program
Iterative? • Should Handle iterative processing like PDE (Partial Differential Equation) • http://www.iterativemapreduce.org/ 7
Graph based Computation • Pregel – Google’s large scale graph • GordenOrb • Giraph – Yahoo’s platform • Hama – Apache Hama’s • Pegasus – Carnegie Melon University 8
Single Source Shortest Path (SSSP)  Problem – Find shortest path from a source node to all target nodes  Solution – MapReduce – Pregel 9
Example: SSSP—using MapReduce • A Map task receives – Key: node n – Value: D (distance from start), points-to (list of nodes reachable from n) – D(n) = dist + min(D(m)) • The Reduce task gathers possible distances and selects the minimum one 10
Example: SSSP—using MapReduce  Adjacency matrix  Adjacency List A: (B, 10), (D, 5) B: (C, 1), (D, 2) C: (E, 4) D: (B, 3), (C, 9), (E, 2) E: (A, 7), (C, 6) A B C D E A 0 10 0 5 0 B 0 0 1 2 0 C 0 0 0 0 4 D 0 3 9 0 2 E 7 0 6 0 0 11 0     10 5 2 3 2 1 9 7 4 6 A B C D E
Example: SSSP—using MapReduce  Map input: <node ID, <dist, adj list>> <A, <0, <(B, 10), (D, 5)>>> <B, <inf, <(C, 1), (D, 2)>>> <C, <inf, <(E, 4)>>> <D, <inf, <(B, 3), (C, 9), (E, 2)>>> <E, <inf, <(A, 7), (C, 6)>>>  Map output: <dest node ID, dist> <B, 10> <D, 5> <C, inf> <D, inf> <E, inf> <B, inf> <C, inf> <E, inf> <A, inf> <C, inf> <A, <0, <(B, 10), (D, 5)>>> <B, <inf, <(C, 1), (D, 2)>>> <C, <inf, <(E, 4)>>> <D, <inf, <(B, 3), (C, 9), (E, 2)>>> <E, <inf, <(A, 7), (C, 6)>>> Flush to local disk! 12 0     10 5 2 3 2 1 9 7 4 6 A B C D E
Example: SSSP—using MapReduce  Reduce input: <node ID, dist> <A, <0, <(B, 10), (D, 5)>>> <A, inf> <B, <inf, <(C, 1), (D, 2)>>> <B, 10> <B, inf> <C, <inf, <(E, 4)>>> <C, inf> <C, inf> <C, inf> <D, <inf, <(B, 3), (C, 9), (E, 2)>>> <D, 5> <D, inf> <E, <inf, <(A, 7), (C, 6)>>> <E, inf> <E, inf> 13 0     10 5 2 3 2 1 9 7 4 6 A B C D E Select possible & minimum value and update former iteration result.
Example: SSSP—using MapReduce  Reduce output: <node ID, <dist, adj list>> = Map input for next iteration <A, <0, <(B, 10), (D, 5)>>> <B, <10, <(C, 1), (D, 2)>>> <C, <inf, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <inf, <(A, 7), (C, 6)>>>  Map output: <dest node ID, dist> <B, 10> <D, 5> <C, 11> <D, 12> <E, inf> <B, 8> <C, 14> <E, 7> <A, inf> <C, inf> <A, <0, <(B, 10), (D, 5)>>> <B, <10, <(C, 1), (D, 2)>>> <C, <inf, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <inf, <(A, 7), (C, 6)>>> Flush to local disk! 14 0 10 5   10 5 2 3 2 1 9 7 4 6 A B C D E
Example: SSSP—using MapReduce  Reduce input: <node ID, dist> <A, <0, <(B, 10), (D, 5)>>> <A, inf> <B, <10, <(C, 1), (D, 2)>>> <B, 10> <B, 8> <C, <inf, <(E, 4)>>> <C, 11> <C, 14> <C, inf> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <D, 5> <D,12> <E, <inf, <(A, 7), (C, 6)>>> <E, inf> <E, 7> 15 0     10 5 2 3 2 1 9 7 4 6 A B C D E Select possible & minimum value and update former iteration result.
Example: SSSP—using MapReduce  Map input: <node ID, <dist, adj list>> <A, <0, <(B, 10), (D, 5)>>> <B, <8, <(C, 1), (D, 2)>>> <C, <11, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <7, <(A, 7), (C, 6)>>>  Map output: <dest node ID, dist> <B, 10> <D, 5> <C, 9> <D, 10> <E, 15> <B, 8> <C, 14> <E, 7> <A, 14> <C, 13> <A, <0, <(B, 10), (D, 5)>>> <B, <8, <(C, 1), (D, 2)>>> <C, <11, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <7, <(A, 7), (C, 6)>>> Flush to local disk! 16 0 8 5 11 7 10 5 2 3 2 1 9 7 4 6 A B C D E
Example: SSSP—using MapReduce  Reduce input: <node ID, dist> <A, <0, <(B, 10), (D, 5)>>> <A, 14> <B, <8, <(C, 1), (D, 2)>>> <B, 10> <B, 8> <C, <11, <(E, 4)>>> <C, 9> <C, 14> <C, 13> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <D, 5> <D, 10> <E, <7, <(A, 7), (C, 6)>>> <E, 15> <E, 7> 17 Select possible & minimum value and update former iteration result. 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 A B C D E
Example: SSSP—using MapReduce  Map input: <node ID, <dist, adj list>> <A, <0, <(B, 10), (D, 5)>>> <B, <8, <(C, 1), (D, 2)>>> <C, <9, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <7, <(A, 7), (C, 6)>>>  Map output: <dest node ID, dist> <B, 10> <D, 5> <C, 9> <D, 10> <E, 13> <B, 8> <C, 14> <E, 7> <A, 14> <C, 13> <A, <0, <(B, 10), (D, 5)>>> <B, <8, <(C, 1), (D, 2)>>> <C, <9, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <7, <(A, 7), (C, 6)>>> Flush to local disk! 18 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 A B C D E
Example: SSSP—using MapReduce  Reduce input: <node ID, dist> <A, <0, <(B, 10), (D, 5)>>> <A, 14> <B, <8, <(C, 1), (D, 2)>>> <B, 10> <B, 8> <C, <9, <(E, 4)>>> <C, 9> <C, 14> <C, 13> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <D, 5><D, 10> <E, <7, <(A, 7), (C, 6)>>> <E, 13> <E, 7> 19 No Changes. Quit Process! 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 A B C D E
• The MapReduce use the key/value pairs to save the nodes and adjacent distance, It is more suitable to process huge datasets rather than the large-scale graph Here, we introduce the new system– Pregel! 20
Model of Computation 21
Model of Pregel Computation Input Output Supersteps: • A sequence of iterations • Vertex compute in parallel Input: a directed graph •Vertex : a vertex ID a modifiable •Edges: a target vertex a modifiable associate with source vertices Output: a directed graph •The set of values explicitly output by the vertices •vertices and edges can be added and moved 22
Maximum Value Example • propagate the largest value to every vertex 23 A B C D
Single Source Shortest Path (SSSP)  Problem – Find shortest path from a source node to all target nodes  Solution – MapReduce – Pregel 24
Example: SSSP—using Pregel 25 0     10 5 2 3 2 1 9 7 4 6 A B D E C
Example: SSSP—using Pregel 26 A ED CB
Example: SSSP—using Pregel 27 0 10 5   10 5 2 3 2 1 9 7 4 6 A ED CB
Example: SSSP—using Pregel 28 A ED CB
Example: SSSP—using Pregel 29 0 8 5 11 7 10 5 2 3 2 1 9 7 4 6 A ED CB
Example: SSSP—using Pregel 30 0 8 5 11 7 10 5 2 3 2 1 9 7 4 6 13 14 15 9 10 A ED CB
Example: SSSP—using Pregel 31 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 A ED CB
Example: SSSP—using Pregel 32 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 13 A ED CB
Example: SSSP—using Pregel 33 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 A ED CB
Pregel vs MapReduce  Pregel – Keeps vertices & edges on the machine that performs computation – Uses network transfers only for messages – Sufficiently expressive, no need for remote reads  MapReduce – Require much more communication and associated overhead – Needs to coordinate the steps of a chained MapReduce add the programming complexity
Pregel Architecture 35
System Architecture  Pregel system uses the master/worker model – Master  Coordinating worker activity  Determines the amount of partitions and assign to worker  Recovers faults of workers (“ping” messges)  Maintains statistics about the progress of computation and the state of the graph – Worker  Maintain the state of its portion of the graph in memory  Executing the Compute() method  Communicates with the other workers 36
37 •Assign portion of the input •Instruct each worker to perform a superstep •call Compute() for each vertex • update the data structure • receive/send messages • responds to master when finished •Control the number of partitions in graph •Notify the master to start the processing
Fault Tolerance  Checkpointing – The master periodically instructs the workers to save the state of their partitions to persistent storage  e.g., Vertex values, edge values, incoming messages  Failure detection – Using regular “ping” messages  Recovery – The master reassigns graph partitions to the currently available workers – The workers all reload their partition state from most recent available checkpoint 38
Goldenorb 39
Goldenorb • Open Source Version of Google’s Pregel • Implemented in Java • Version 0.1.1 • Requirements - hadoop file system - zookeeper for communication 40
41 Orbcluster JobsInProgr ess Jobid Messages heartbeat OrbTracker LeaderGrou p JobQue OrbTrackte rs ZK-TREE Orb-Tracker(L) Orb-Tracker(S) Orb-Tracker(S) Job manager Partition manager Watcher /Event Partition request Partition (Master) … Partition (slave) Partition (slave) Inbound msg queue outbound msg queue current msg queue Inbound msg queue outbound msg queue current msg queue •startLoadVerticesBarrier •Superstep start Barrier •doneComputingVerticesBarr •doneSendingMessageBarrier … •LeaderShipChange •LostMember •NewMember •JobStart/Death/Complete… HDFS Read/write
Message Exchange • Message교환은 Superstep간에 이루어짐 • [S-1] superstep의 outbound message들은 [s] superstep의 inbound messages • Outbound Queue가 가득차면 message들을 보내고 다시 queuing • Superstep 중간에 message를 받은 partition은 inbound queue에 저장하고 다음 Superstep까지 보관 • 현재 superstep에 사용할 message들은 current message queue에 복사 • 이 때, inbound queue가 system이나 jvm의 memory size 를 넘어서면 overflow 발생
Memory management • Outbound Message Queue - Fixed size, 가득 차면 바로 messages 보냄 • Inbound Message Queue - 다음 Superstep에 사용 - Message 양이 많아지면 overflow가능성 있음 • Current Message Queue - Inbound Queue 과 같은 사이즈 - 현 Superstep 에서는 CurrentQueue에 inboundQueue를 복사해서 사용하므로 currentQueue+inboundQueue 의 메모리 사용 overflow  Inbound Queue를 file 기반의 local 저장공간에 구현 필요
API • Sub-classing the predefined classes – Reader/writer/vertex/message 44 Class Vertex { public Vertex(Class<VV> vertexValue, Class<EV> edgeValue, Class<MV> messageValue); String vertexID(); abstract void compute(Collection<MV> messages); long superStep(); void setValue(VV value); VV getValue(); Collection<Edge<EV>> getEdges(); void sendMessage(MV message); void voteToHalt(); }
Not yet implemented • Aggregator – a mechanism for global communication monitoring and data • Combiner – Reducing the number of messages – Ex) if compute() sum messages’ value, combiner can calculate and transmit single message(sum) • Topology mutation – Remove or add Vertex/Edge • Fault Recovery 45
Implementation 46
Implement  Maximum Value  Single Source Shortest Path  PageRank  K-means  Mean-shift 47
48 MaximumValue
PageRank • PageRank is Google’s way of deciding a Page’s importance • A important page is linked to by many pages with high PageRank • PR(A) = PR(inLink_v1)/L(t1) + ….+ P(inLink_vn)/L(tn) • Add damping factor d • PR(A) = (1-d) + d∑PR(v)/L(v) • Repeat until converged 49
PageRank 50
PageRank 51 AE B C D F <Input file> <output file>
K-means • N observations are parted to k cluster • Each observation belongs to the cluster with the nearest mean No object move group? End Number of cluster K Calculate centroids Distance objects to centroids Grouping based on minimum distance start NO YES
K-means 53 • Message includes cluster id and value • Every superstep, a vertex sends message to all vertices 1 2 3 100 101 102 seed2 seed1 A B C D E F Step A B C D E F S0 C1 C2 - - - -
K-means 54 1 2 3 100 101 102 A B C D E F Step A B C D E F S0 C1 C2 - - - - S1 C1 C2 C1 C1 C1 C1 Centroid1 = Value(A) Centroid2 = Value(B) 1 2 3 100 101 102 seed2 seed1 A B C D E F
K-means 55 Step A B C D E F S0 C1 C2 - - - - S1 C1 C2 C1 C1 C1 C1 S2 C2 C2 C2 C1 C1 C1 1 2 3 100 101 102 A B C D E F Centroid1 = Mean(A,C,D,E,F) Centroid2 = Mean(B) 1 2 3 100 101 102 A B C D E F
K-means 56 Step A B C D E F S0 C1 C2 - - - - S1 C1 C2 C1 C1 C1 C1 S2 C2 C2 C2 C1 C1 C1 S3 C2 C2 C2 C1 C1 C1 1 2 3 100 101 102 A B C D E F Centroid1 = Mean(D,E,F) Centroid2 = Mean(A,B,C) If centroids are converged, Quit Process!
57
K-means 58 <input file> <output file> N : number of vertices Each superstep, NxN messages are exchanged.  O(n2) : need too much memory !!!
Giraph 59
Giraph • ASF(Apache Software Foundation)’s Open Source Version of Google’s Pregel • Implemented in Java • Apache incubator • Requirements - hadoop 0.20.203 or higher version : map-only job in hadoop - zookeeper : if not exist, use hadoop file system instead of zookeeper 60
Giraph – vertex distribution 61
Giraph - usages • Users can set the checkpoint frequency – GiraphJob.getConfiguration().set(“giraph.checkpointFrequency”, 0) //means no check points • User should set zookeeper configuration – GiraphJob.setZookeeperConfiguration(“zk-server-list”); 62
Giraph - Characteristics • Faulty tolerance – If the master dies, a new one will automatically take over – If a worker dies, the app is rolled back to a previously checkpointed superstep – If a zookeeper server dies, as long as a quorum remains, the app can proceed – But, Hadoop SPOF still exist • Combiner/Aggregator • JSON in/out format • Easy Job status monitoring (http) 63
Experiments 64
Experiments • 3 severs • nPartition = nMapper = 9 • MR vs GoldenOrb vs Giraph – PageRank – Kmeans (mahout) – Elapse time, cpu, memory, disk, network 65
Experiments - PageRank • Number of Vertices ≈ 220,000 • Fixed iteration = 100 66 Elapse Time CPU (%) Memory (kb) Network (bytes) Disk Write (sec/s) Rcv. Trans. read write GoldenOrb 1m 56s 14.53 3,745,376 19,437 12,845 777 606 Giraph 3m 31s 8.77 1,244,000 11,374 914 0 326 MapReduce 34m 51s 3.75 3,091,239 13,514 867 0 4101
Experiments - Kmeans • Number of Vertices = 100,000 • Number of Cluster(K) = 10 67 Elapse Time CPU (%) Memory (kb) Network (bytes) Disk Write (sec/s) Rcv. Trans. read write GoldenOrb 3m 19s 13.32 3,857,892 11,634 27,086 128 1151 Giraph 1m 49s 6.36 1,245,000 7,810 1,999 0 536 MapReduce 11m 28s 1.48 2,645,517 13,528 1,005 0 7104
Experiments 68 Elapse Time(s)
Installation 69
Install Goldenorb (1) • Requirements - hadoop-0.20.2 - zookeeper-3.3.3 • Download & unzip – org.goldenorb.core-0.1.1-SNAPSHOT-distribution.zip 70
Install Goldenorb(2) • Set configuration ① ORB_HOME 환경변수 > export ORB_HOME=/usr/local/goldenorb ② Conf/orbServers > localhost:/usr/local/goldenorb ③ Conf/orb-site.xml > cp orb-site.sample.xml orb-site.xml > vi orb-site.xml ④ If Distributed mode , copy to all servers 71 <property> <name>goldenOrb.zookeeper.quorum</name> <value> localhost</value> <description> The server running zookeeper</description> <property> …… <property> <name>fs.default.name</name> <value>hdfs://localhost:9000</value> </property> Target zookeeper server IP
Install Goldenorb(3) • Set running environment ① Hadoop 실행 > $HADOOP_HOME/bin/start-dfs.sh ② Zookeeper 실행 > $ZK_HOME/bin/zkServer.sh start ③ Orb-tracker 실행 > $ORB_HOME/bin/orb-tracker.sh start ④ Log 확인 > cat #ORB_HOME/logs/xxx.log 72
Install Goldenorb(4) • Make input - ex) maximum value < Vertex-id > <value> <outgoing-edge-list> 73 0 8 5 11 7 A B C D E A 0 B D B 8 C D C 11 E D 5 B C E E 7 A C
Install Goldenorb(5) • Upload input files > hadoop fs –put maxvalue.txt /test/ • Run > java -cp conf/.:org.goldenorb.core-0.1.1-SNAPSHOT.jar:lib/*:yourjar.jar org.goldenOrb.algorithms.YourAlgorithm goldenOrb.orb.localFilesToDistribute=/home/user/yourjar.jar mapred.input.dir=/test/maxvaluetxt/ mapred.output.dir=/test/output goldenOrb.orb.requestedPartitions=3 goldenOrb.orb.reservedPartitions=0 goldenOrb.orb.classpaths=yourjar.jar • Result > hadoop fs –ls /test/output > hadoop fs –cat /test/output/* 74
Install Giraph(1) • Requirements - hadoop-0.20.203 - zookeeper-3.3.3 - maven 3.0.3 • Download > svn checkout http://svn.apache.org/repos/asf/incubator/giraph/trunk giraph • Compile > mvn compile – Generate target/giraph-{version}-jar-with-dependencies.jar 75
Install Giraph(2) • Set running environment ① >$HADOOP_HOME/bin/start-all.sh ② >$ZK_HOME/zkServer.sh start • Upload input file to HDFS > hadoop fs –put test.grf /giraph/test/input/ * Input Format (JASON) 76 [0,0,[[1,0]]] [1,0,[[2,100]]] [2,100,[[3,200]]] [3,300,[[4,300]]] [4,600,[[5,400]]]
Install Giraph(3) • Run - ex)shortest path algorithm > hadoop jar giraph-{version}-jar-with-dependencies.jar org.apache.giraph.examples.SimpleShortestPathsVertex <input-path> <output-path> <source-node> <number of worker> • Running status – http://localhost:50030 • Result > hadoop fs –cat /<output-path/part-* 77
Orb vs Giraph 78 GoldenOrb Giraph 상태 확인 로그 파일 용이 Fault Tolerance X O Vertex mutation, Combiner, Aggregator… X O 개발환경 X O I/O format O X Update X O
Thank you ! 79

More Related Content

PDF
Gate Previous Years Papers
byRahul Jain
 
PDF
Efficient computational modelling for integrated power plant design - Prof Er...
byUK Carbon Capture and Storage Research Centre
 
PDF
2009 : Solving linear optimization problems with MOSEK
byjensenbo
 
PPT
Basics of programming embedded processors
byMurphy Chen
 
PDF
ANPR FPGA Workshop
bySundance Multiprocessor Technology Ltd.
 
PDF
OPTEX Mathematical Modeling and Management System
byJesus Velasquez
 
DOCX
Arna Friend Controls II Final
byArna Friend
 
PDF
UK ATC 2015: Automated Post Processing of Multimodel Optimisation Data
byAltair
 
Gate Previous Years Papers
byRahul Jain
 
Efficient computational modelling for integrated power plant design - Prof Er...
byUK Carbon Capture and Storage Research Centre
 
2009 : Solving linear optimization problems with MOSEK
byjensenbo
 
Basics of programming embedded processors
byMurphy Chen
 
ANPR FPGA Workshop
bySundance Multiprocessor Technology Ltd.
 
OPTEX Mathematical Modeling and Management System
byJesus Velasquez
 
Arna Friend Controls II Final
byArna Friend
 
UK ATC 2015: Automated Post Processing of Multimodel Optimisation Data
byAltair
 

Similar to Graph analysis platform comparison, pregel/goldenorb/giraph

ODP
Processing Large Graphs in Hadoop
byDani Solà Lagares
 
PPTX
2011.10.14 Apache Giraph - Hortonworks
byAvery Ching
 
PDF
Simplified Data Processing On Large Cluster
byHarsh Kevadia
 
PDF
Parallel Data Processing with MapReduce: A Survey
byKyong-Ha Lee
 
PPTX
Pregel
byWeiru Dai
 
PPTX
Cloud computing_processing frameworks
byReem Abdel-Rahman
 
PDF
Introducing Apache Giraph for Large Scale Graph Processing
bysscdotopen
 
PPT
Lecture Slide - Introduction to Hadoop, HDFS, MapR.ppt
bySuchithraaPalani
 
PDF
Hadoop Overview kdd2011
byMilind Bhandarkar
 
PDF
Distributed Computing with Apache Hadoop. Introduction to MapReduce.
byKonstantin V. Shvachko
 
PDF
Hadoop Overview & Architecture
byEMC
 
PPTX
Automatic Scaling Iterative Computations
byGuozhang Wang
 
PPTX
This gives a brief detail about big data
bychinky1118
 
PPTX
Scalable Distributed Graph Algorithms on Apache Spark
byLynxAnalytics
 
PDF
Map reduce
byxydii
 
PDF
Map reduceoriginalpaper mandatoryreading
bycoolmirza143
 
PPTX
ch02-mapreduce.pptx
byGiannisPagges
 
PPTX
Introduction to map reduce
byM Baddar
 
PDF
Hadoop map reduce concepts
bySubhas Kumar Ghosh
 
PDF
Thorny Path to the Large Scale Graph Processing, Алексей Зиновьев (Тамтэк)
byOntico
 
Processing Large Graphs in Hadoop
byDani Solà Lagares
 
2011.10.14 Apache Giraph - Hortonworks
byAvery Ching
 
Simplified Data Processing On Large Cluster
byHarsh Kevadia
 
Parallel Data Processing with MapReduce: A Survey
byKyong-Ha Lee
 
Pregel
byWeiru Dai
 
Cloud computing_processing frameworks
byReem Abdel-Rahman
 
Introducing Apache Giraph for Large Scale Graph Processing
bysscdotopen
 
Lecture Slide - Introduction to Hadoop, HDFS, MapR.ppt
bySuchithraaPalani
 
Hadoop Overview kdd2011
byMilind Bhandarkar
 
Distributed Computing with Apache Hadoop. Introduction to MapReduce.
byKonstantin V. Shvachko
 
Hadoop Overview & Architecture
byEMC
 
Automatic Scaling Iterative Computations
byGuozhang Wang
 
This gives a brief detail about big data
bychinky1118
 
Scalable Distributed Graph Algorithms on Apache Spark
byLynxAnalytics
 
Map reduce
byxydii
 
Map reduceoriginalpaper mandatoryreading
bycoolmirza143
 
ch02-mapreduce.pptx
byGiannisPagges
 
Introduction to map reduce
byM Baddar
 
Hadoop map reduce concepts
bySubhas Kumar Ghosh
 
Thorny Path to the Large Scale Graph Processing, Алексей Зиновьев (Тамтэк)
byOntico
 

More from Andrew Yongjoon Kong

PDF
Tunnel without tunnel
byAndrew Yongjoon Kong
 
PDF
Nightmare with ceph : Recovery from ceph cluster total failure
byAndrew Yongjoon Kong
 
POTX
Stream analysis with kafka native way and considerations about monitoring as ...
byAndrew Yongjoon Kong
 
PDF
Automating auto-scaled load balancer based on linux and vm orchestrator
byAndrew Yongjoon Kong
 
PDF
GPU cloud with Job scheduler and Container
byAndrew Yongjoon Kong
 
PDF
Embracing clouds
byAndrew Yongjoon Kong
 
PDF
openstack, devops and people
byAndrew Yongjoon Kong
 
PDF
Cloud data center and openstack
byAndrew Yongjoon Kong
 
PDF
Openstack summit 2015
byAndrew Yongjoon Kong
 
PDF
Cloud: From Unmanned Data Center to Algorithmic Economy using Openstack
byAndrew Yongjoon Kong
 
PDF
Way to cloud
byAndrew Yongjoon Kong
 
PPTX
Openstack dev on
byAndrew Yongjoon Kong
 
Tunnel without tunnel
byAndrew Yongjoon Kong
 
Nightmare with ceph : Recovery from ceph cluster total failure
byAndrew Yongjoon Kong
 
Stream analysis with kafka native way and considerations about monitoring as ...
byAndrew Yongjoon Kong
 
Automating auto-scaled load balancer based on linux and vm orchestrator
byAndrew Yongjoon Kong
 
GPU cloud with Job scheduler and Container
byAndrew Yongjoon Kong
 
Embracing clouds
byAndrew Yongjoon Kong
 
openstack, devops and people
byAndrew Yongjoon Kong
 
Cloud data center and openstack
byAndrew Yongjoon Kong
 
Openstack summit 2015
byAndrew Yongjoon Kong
 
Cloud: From Unmanned Data Center to Algorithmic Economy using Openstack
byAndrew Yongjoon Kong
 
Way to cloud
byAndrew Yongjoon Kong
 
Openstack dev on
byAndrew Yongjoon Kong
 

Recently uploaded

PDF
GTI Solution Overview.pdf useful for security solution
by33fastfast
 
PPTX
Social_Media_Good_or_Bad_B2Plus_Lesson.pptx
byJuliaRutkowska4
 
PDF
Batman Forever album sličice.pdf panini album
byStefanFilipovic9
 
PPTX
Creating content that converts into leads.
bySEONutri
 
PDF
automobili.pdf panini album slicice ......
byStefanFilipovic9
 
PDF
investor pitch - iapploud | independent music platform
byVenkat Subramanian
 
GTI Solution Overview.pdf useful for security solution
by33fastfast
 
Social_Media_Good_or_Bad_B2Plus_Lesson.pptx
byJuliaRutkowska4
 
Batman Forever album sličice.pdf panini album
byStefanFilipovic9
 
Creating content that converts into leads.
bySEONutri
 
automobili.pdf panini album slicice ......
byStefanFilipovic9
 
investor pitch - iapploud | independent music platform
byVenkat Subramanian
 

Graph analysis platform comparison, pregel/goldenorb/giraph

  • 1.
    A System forLarge-Scale Graph Processing Pregel, GoldenOrb, Giraph 2012-07-18 Andrew Yongjoon Kong sstrato.open@gmail.com 1
  • 2.
    Contents • Introduction • Modelof Computation • Pregel Architecture • Goldenorb • Implementation • Future work 2
  • 3.
  • 4.
    Introduction • Today, Manypractical computing problems concern large graphs • Applied algorithms - Shortest paths computations - Page rank - Clustering techniques • MapReduce is ill-suited for graph processing - Many iterations are needed for parallel graph processing - Materializations of intermediate results at every MapReduce iteration harm performance 4
  • 5.
    Introduction • Hadoop iswell-suited for non-iterative, data parallelized processing 5 Smith Waterman is a non iterativ e case and of c ourse runs fine
  • 6.
    Introduction 6 map map reduce Compute thedist ance to each dat a point from eac h cluster center a nd assign points to cluster centers Compute new cluster centers Compute new clust er centers User program
  • 7.
    Iterative? • Should Handleiterative processing like PDE (Partial Differential Equation) • http://www.iterativemapreduce.org/ 7
  • 8.
    Graph based Computation •Pregel – Google’s large scale graph • GordenOrb • Giraph – Yahoo’s platform • Hama – Apache Hama’s • Pegasus – Carnegie Melon University 8
  • 9.
    Single Source ShortestPath (SSSP)  Problem – Find shortest path from a source node to all target nodes  Solution – MapReduce – Pregel 9
  • 10.
    Example: SSSP—using MapReduce •A Map task receives – Key: node n – Value: D (distance from start), points-to (list of nodes reachable from n) – D(n) = dist + min(D(m)) • The Reduce task gathers possible distances and selects the minimum one 10
  • 11.
    Example: SSSP—using MapReduce Adjacency matrix  Adjacency List A: (B, 10), (D, 5) B: (C, 1), (D, 2) C: (E, 4) D: (B, 3), (C, 9), (E, 2) E: (A, 7), (C, 6) A B C D E A 0 10 0 5 0 B 0 0 1 2 0 C 0 0 0 0 4 D 0 3 9 0 2 E 7 0 6 0 0 11 0     10 5 2 3 2 1 9 7 4 6 A B C D E
  • 12.
    Example: SSSP—using MapReduce Map input: <node ID, <dist, adj list>> <A, <0, <(B, 10), (D, 5)>>> <B, <inf, <(C, 1), (D, 2)>>> <C, <inf, <(E, 4)>>> <D, <inf, <(B, 3), (C, 9), (E, 2)>>> <E, <inf, <(A, 7), (C, 6)>>>  Map output: <dest node ID, dist> <B, 10> <D, 5> <C, inf> <D, inf> <E, inf> <B, inf> <C, inf> <E, inf> <A, inf> <C, inf> <A, <0, <(B, 10), (D, 5)>>> <B, <inf, <(C, 1), (D, 2)>>> <C, <inf, <(E, 4)>>> <D, <inf, <(B, 3), (C, 9), (E, 2)>>> <E, <inf, <(A, 7), (C, 6)>>> Flush to local disk! 12 0     10 5 2 3 2 1 9 7 4 6 A B C D E
  • 13.
    Example: SSSP—using MapReduce Reduce input: <node ID, dist> <A, <0, <(B, 10), (D, 5)>>> <A, inf> <B, <inf, <(C, 1), (D, 2)>>> <B, 10> <B, inf> <C, <inf, <(E, 4)>>> <C, inf> <C, inf> <C, inf> <D, <inf, <(B, 3), (C, 9), (E, 2)>>> <D, 5> <D, inf> <E, <inf, <(A, 7), (C, 6)>>> <E, inf> <E, inf> 13 0     10 5 2 3 2 1 9 7 4 6 A B C D E Select possible & minimum value and update former iteration result.
  • 14.
    Example: SSSP—using MapReduce Reduce output: <node ID, <dist, adj list>> = Map input for next iteration <A, <0, <(B, 10), (D, 5)>>> <B, <10, <(C, 1), (D, 2)>>> <C, <inf, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <inf, <(A, 7), (C, 6)>>>  Map output: <dest node ID, dist> <B, 10> <D, 5> <C, 11> <D, 12> <E, inf> <B, 8> <C, 14> <E, 7> <A, inf> <C, inf> <A, <0, <(B, 10), (D, 5)>>> <B, <10, <(C, 1), (D, 2)>>> <C, <inf, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <inf, <(A, 7), (C, 6)>>> Flush to local disk! 14 0 10 5   10 5 2 3 2 1 9 7 4 6 A B C D E
  • 15.
    Example: SSSP—using MapReduce Reduce input: <node ID, dist> <A, <0, <(B, 10), (D, 5)>>> <A, inf> <B, <10, <(C, 1), (D, 2)>>> <B, 10> <B, 8> <C, <inf, <(E, 4)>>> <C, 11> <C, 14> <C, inf> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <D, 5> <D,12> <E, <inf, <(A, 7), (C, 6)>>> <E, inf> <E, 7> 15 0     10 5 2 3 2 1 9 7 4 6 A B C D E Select possible & minimum value and update former iteration result.
  • 16.
    Example: SSSP—using MapReduce Map input: <node ID, <dist, adj list>> <A, <0, <(B, 10), (D, 5)>>> <B, <8, <(C, 1), (D, 2)>>> <C, <11, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <7, <(A, 7), (C, 6)>>>  Map output: <dest node ID, dist> <B, 10> <D, 5> <C, 9> <D, 10> <E, 15> <B, 8> <C, 14> <E, 7> <A, 14> <C, 13> <A, <0, <(B, 10), (D, 5)>>> <B, <8, <(C, 1), (D, 2)>>> <C, <11, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <7, <(A, 7), (C, 6)>>> Flush to local disk! 16 0 8 5 11 7 10 5 2 3 2 1 9 7 4 6 A B C D E
  • 17.
    Example: SSSP—using MapReduce Reduce input: <node ID, dist> <A, <0, <(B, 10), (D, 5)>>> <A, 14> <B, <8, <(C, 1), (D, 2)>>> <B, 10> <B, 8> <C, <11, <(E, 4)>>> <C, 9> <C, 14> <C, 13> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <D, 5> <D, 10> <E, <7, <(A, 7), (C, 6)>>> <E, 15> <E, 7> 17 Select possible & minimum value and update former iteration result. 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 A B C D E
  • 18.
    Example: SSSP—using MapReduce Map input: <node ID, <dist, adj list>> <A, <0, <(B, 10), (D, 5)>>> <B, <8, <(C, 1), (D, 2)>>> <C, <9, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <7, <(A, 7), (C, 6)>>>  Map output: <dest node ID, dist> <B, 10> <D, 5> <C, 9> <D, 10> <E, 13> <B, 8> <C, 14> <E, 7> <A, 14> <C, 13> <A, <0, <(B, 10), (D, 5)>>> <B, <8, <(C, 1), (D, 2)>>> <C, <9, <(E, 4)>>> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <E, <7, <(A, 7), (C, 6)>>> Flush to local disk! 18 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 A B C D E
  • 19.
    Example: SSSP—using MapReduce Reduce input: <node ID, dist> <A, <0, <(B, 10), (D, 5)>>> <A, 14> <B, <8, <(C, 1), (D, 2)>>> <B, 10> <B, 8> <C, <9, <(E, 4)>>> <C, 9> <C, 14> <C, 13> <D, <5, <(B, 3), (C, 9), (E, 2)>>> <D, 5><D, 10> <E, <7, <(A, 7), (C, 6)>>> <E, 13> <E, 7> 19 No Changes. Quit Process! 0 8 5 9 7 10 5 2 3 2 1 9 7 4 6 A B C D E
  • 20.
    • The MapReduceuse the key/value pairs to save the nodes and adjacent distance, It is more suitable to process huge datasets rather than the large-scale graph Here, we introduce the new system– Pregel! 20
  • 21.
  • 22.
    Model of PregelComputation Input Output Supersteps: • A sequence of iterations • Vertex compute in parallel Input: a directed graph •Vertex : a vertex ID a modifiable •Edges: a target vertex a modifiable associate with source vertices Output: a directed graph •The set of values explicitly output by the vertices •vertices and edges can be added and moved 22
  • 23.
    Maximum Value Example •propagate the largest value to every vertex 23 A B C D
  • 24.
    Single Source ShortestPath (SSSP)  Problem – Find shortest path from a source node to all target nodes  Solution – MapReduce – Pregel 24
  • 25.
  • 26.
  • 27.
  • 28.
  • 29.
  • 30.
    Example: SSSP—using Pregel 30 0 8 5 11 7 10 5 23 2 1 9 7 4 6 13 14 15 9 10 A ED CB
  • 31.
  • 32.
  • 33.
  • 34.
    Pregel vs MapReduce Pregel – Keeps vertices & edges on the machine that performs computation – Uses network transfers only for messages – Sufficiently expressive, no need for remote reads  MapReduce – Require much more communication and associated overhead – Needs to coordinate the steps of a chained MapReduce add the programming complexity
  • 35.
  • 36.
    System Architecture  Pregelsystem uses the master/worker model – Master  Coordinating worker activity  Determines the amount of partitions and assign to worker  Recovers faults of workers (“ping” messges)  Maintains statistics about the progress of computation and the state of the graph – Worker  Maintain the state of its portion of the graph in memory  Executing the Compute() method  Communicates with the other workers 36
  • 37.
    37 •Assign portion ofthe input •Instruct each worker to perform a superstep •call Compute() for each vertex • update the data structure • receive/send messages • responds to master when finished •Control the number of partitions in graph •Notify the master to start the processing
  • 38.
    Fault Tolerance  Checkpointing –The master periodically instructs the workers to save the state of their partitions to persistent storage  e.g., Vertex values, edge values, incoming messages  Failure detection – Using regular “ping” messages  Recovery – The master reassigns graph partitions to the currently available workers – The workers all reload their partition state from most recent available checkpoint 38
  • 39.
  • 40.
    Goldenorb • Open SourceVersion of Google’s Pregel • Implemented in Java • Version 0.1.1 • Requirements - hadoop file system - zookeeper for communication 40
  • 41.
    41 Orbcluster JobsInProgr ess Jobid Messages heartbeat OrbTracker LeaderGrou p JobQue OrbTrackte rs ZK-TREE Orb-Tracker(L) Orb-Tracker(S) Orb-Tracker(S) Job manager Partition manager Watcher /Event Partition request Partition (Master) … Partition (slave) Partition (slave) Inbound msgqueue outbound msg queue current msg queue Inbound msg queue outbound msg queue current msg queue •startLoadVerticesBarrier •Superstep start Barrier •doneComputingVerticesBarr •doneSendingMessageBarrier … •LeaderShipChange •LostMember •NewMember •JobStart/Death/Complete… HDFS Read/write
  • 42.
    Message Exchange • Message교환은Superstep간에 이루어짐 • [S-1] superstep의 outbound message들은 [s] superstep의 inbound messages • Outbound Queue가 가득차면 message들을 보내고 다시 queuing • Superstep 중간에 message를 받은 partition은 inbound queue에 저장하고 다음 Superstep까지 보관 • 현재 superstep에 사용할 message들은 current message queue에 복사 • 이 때, inbound queue가 system이나 jvm의 memory size 를 넘어서면 overflow 발생
  • 43.
    Memory management • OutboundMessage Queue - Fixed size, 가득 차면 바로 messages 보냄 • Inbound Message Queue - 다음 Superstep에 사용 - Message 양이 많아지면 overflow가능성 있음 • Current Message Queue - Inbound Queue 과 같은 사이즈 - 현 Superstep 에서는 CurrentQueue에 inboundQueue를 복사해서 사용하므로 currentQueue+inboundQueue 의 메모리 사용 overflow  Inbound Queue를 file 기반의 local 저장공간에 구현 필요
  • 44.
    API • Sub-classing thepredefined classes – Reader/writer/vertex/message 44 Class Vertex { public Vertex(Class<VV> vertexValue, Class<EV> edgeValue, Class<MV> messageValue); String vertexID(); abstract void compute(Collection<MV> messages); long superStep(); void setValue(VV value); VV getValue(); Collection<Edge<EV>> getEdges(); void sendMessage(MV message); void voteToHalt(); }
  • 45.
    Not yet implemented •Aggregator – a mechanism for global communication monitoring and data • Combiner – Reducing the number of messages – Ex) if compute() sum messages’ value, combiner can calculate and transmit single message(sum) • Topology mutation – Remove or add Vertex/Edge • Fault Recovery 45
  • 46.
  • 47.
    Implement  Maximum Value Single Source Shortest Path  PageRank  K-means  Mean-shift 47
  • 48.
  • 49.
    PageRank • PageRank isGoogle’s way of deciding a Page’s importance • A important page is linked to by many pages with high PageRank • PR(A) = PR(inLink_v1)/L(t1) + ….+ P(inLink_vn)/L(tn) • Add damping factor d • PR(A) = (1-d) + d∑PR(v)/L(v) • Repeat until converged 49
  • 50.
  • 51.
  • 52.
    K-means • N observationsare parted to k cluster • Each observation belongs to the cluster with the nearest mean No object move group? End Number of cluster K Calculate centroids Distance objects to centroids Grouping based on minimum distance start NO YES
  • 53.
    K-means 53 • Message includescluster id and value • Every superstep, a vertex sends message to all vertices 1 2 3 100 101 102 seed2 seed1 A B C D E F Step A B C D E F S0 C1 C2 - - - -
  • 54.
    K-means 54 1 2 3 100 101 102 A B C D E F Step AB C D E F S0 C1 C2 - - - - S1 C1 C2 C1 C1 C1 C1 Centroid1 = Value(A) Centroid2 = Value(B) 1 2 3 100 101 102 seed2 seed1 A B C D E F
  • 55.
    K-means 55 Step A BC D E F S0 C1 C2 - - - - S1 C1 C2 C1 C1 C1 C1 S2 C2 C2 C2 C1 C1 C1 1 2 3 100 101 102 A B C D E F Centroid1 = Mean(A,C,D,E,F) Centroid2 = Mean(B) 1 2 3 100 101 102 A B C D E F
  • 56.
    K-means 56 Step A BC D E F S0 C1 C2 - - - - S1 C1 C2 C1 C1 C1 C1 S2 C2 C2 C2 C1 C1 C1 S3 C2 C2 C2 C1 C1 C1 1 2 3 100 101 102 A B C D E F Centroid1 = Mean(D,E,F) Centroid2 = Mean(A,B,C) If centroids are converged, Quit Process!
  • 57.
  • 58.
    K-means 58 <input file> <output file> N: number of vertices Each superstep, NxN messages are exchanged.  O(n2) : need too much memory !!!
  • 59.
  • 60.
    Giraph • ASF(Apache SoftwareFoundation)’s Open Source Version of Google’s Pregel • Implemented in Java • Apache incubator • Requirements - hadoop 0.20.203 or higher version : map-only job in hadoop - zookeeper : if not exist, use hadoop file system instead of zookeeper 60
  • 61.
    Giraph – vertexdistribution 61
  • 62.
    Giraph - usages •Users can set the checkpoint frequency – GiraphJob.getConfiguration().set(“giraph.checkpointFrequency”, 0) //means no check points • User should set zookeeper configuration – GiraphJob.setZookeeperConfiguration(“zk-server-list”); 62
  • 63.
    Giraph - Characteristics •Faulty tolerance – If the master dies, a new one will automatically take over – If a worker dies, the app is rolled back to a previously checkpointed superstep – If a zookeeper server dies, as long as a quorum remains, the app can proceed – But, Hadoop SPOF still exist • Combiner/Aggregator • JSON in/out format • Easy Job status monitoring (http) 63
  • 64.
  • 65.
    Experiments • 3 severs •nPartition = nMapper = 9 • MR vs GoldenOrb vs Giraph – PageRank – Kmeans (mahout) – Elapse time, cpu, memory, disk, network 65
  • 66.
    Experiments - PageRank •Number of Vertices ≈ 220,000 • Fixed iteration = 100 66 Elapse Time CPU (%) Memory (kb) Network (bytes) Disk Write (sec/s) Rcv. Trans. read write GoldenOrb 1m 56s 14.53 3,745,376 19,437 12,845 777 606 Giraph 3m 31s 8.77 1,244,000 11,374 914 0 326 MapReduce 34m 51s 3.75 3,091,239 13,514 867 0 4101
  • 67.
    Experiments - Kmeans •Number of Vertices = 100,000 • Number of Cluster(K) = 10 67 Elapse Time CPU (%) Memory (kb) Network (bytes) Disk Write (sec/s) Rcv. Trans. read write GoldenOrb 3m 19s 13.32 3,857,892 11,634 27,086 128 1151 Giraph 1m 49s 6.36 1,245,000 7,810 1,999 0 536 MapReduce 11m 28s 1.48 2,645,517 13,528 1,005 0 7104
  • 68.
  • 69.
  • 70.
    Install Goldenorb (1) •Requirements - hadoop-0.20.2 - zookeeper-3.3.3 • Download & unzip – org.goldenorb.core-0.1.1-SNAPSHOT-distribution.zip 70
  • 71.
    Install Goldenorb(2) • Setconfiguration ① ORB_HOME 환경변수 > export ORB_HOME=/usr/local/goldenorb ② Conf/orbServers > localhost:/usr/local/goldenorb ③ Conf/orb-site.xml > cp orb-site.sample.xml orb-site.xml > vi orb-site.xml ④ If Distributed mode , copy to all servers 71 <property> <name>goldenOrb.zookeeper.quorum</name> <value> localhost</value> <description> The server running zookeeper</description> <property> …… <property> <name>fs.default.name</name> <value>hdfs://localhost:9000</value> </property> Target zookeeper server IP
  • 72.
    Install Goldenorb(3) • Setrunning environment ① Hadoop 실행 > $HADOOP_HOME/bin/start-dfs.sh ② Zookeeper 실행 > $ZK_HOME/bin/zkServer.sh start ③ Orb-tracker 실행 > $ORB_HOME/bin/orb-tracker.sh start ④ Log 확인 > cat #ORB_HOME/logs/xxx.log 72
  • 73.
    Install Goldenorb(4) • Makeinput - ex) maximum value < Vertex-id > <value> <outgoing-edge-list> 73 0 8 5 11 7 A B C D E A 0 B D B 8 C D C 11 E D 5 B C E E 7 A C
  • 74.
    Install Goldenorb(5) • Uploadinput files > hadoop fs –put maxvalue.txt /test/ • Run > java -cp conf/.:org.goldenorb.core-0.1.1-SNAPSHOT.jar:lib/*:yourjar.jar org.goldenOrb.algorithms.YourAlgorithm goldenOrb.orb.localFilesToDistribute=/home/user/yourjar.jar mapred.input.dir=/test/maxvaluetxt/ mapred.output.dir=/test/output goldenOrb.orb.requestedPartitions=3 goldenOrb.orb.reservedPartitions=0 goldenOrb.orb.classpaths=yourjar.jar • Result > hadoop fs –ls /test/output > hadoop fs –cat /test/output/* 74
  • 75.
    Install Giraph(1) • Requirements -hadoop-0.20.203 - zookeeper-3.3.3 - maven 3.0.3 • Download > svn checkout http://svn.apache.org/repos/asf/incubator/giraph/trunk giraph • Compile > mvn compile – Generate target/giraph-{version}-jar-with-dependencies.jar 75
  • 76.
    Install Giraph(2) • Setrunning environment ① >$HADOOP_HOME/bin/start-all.sh ② >$ZK_HOME/zkServer.sh start • Upload input file to HDFS > hadoop fs –put test.grf /giraph/test/input/ * Input Format (JASON) 76 [0,0,[[1,0]]] [1,0,[[2,100]]] [2,100,[[3,200]]] [3,300,[[4,300]]] [4,600,[[5,400]]]
  • 77.
    Install Giraph(3) • Run -ex)shortest path algorithm > hadoop jar giraph-{version}-jar-with-dependencies.jar org.apache.giraph.examples.SimpleShortestPathsVertex <input-path> <output-path> <source-node> <number of worker> • Running status – http://localhost:50030 • Result > hadoop fs –cat /<output-path/part-* 77
  • 78.
    Orb vs Giraph 78 GoldenOrbGiraph 상태 확인 로그 파일 용이 Fault Tolerance X O Vertex mutation, Combiner, Aggregator… X O 개발환경 X O I/O format O X Update X O
  • 79.