![Page 2: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/2.jpg)
Outlines
• Motivation• Introduction• Evaluation• Conclusion• Questions
![Page 3: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/3.jpg)
Motivation
• A credit card size Raspberry Pi can run general Linux with very low power consumption
![Page 4: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/4.jpg)
Motivation
• ARM cluster vs. x86_64 cluster
![Page 5: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/5.jpg)
Introduction
• Hadoop MapReduce• Cubieboard2
![Page 6: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/6.jpg)
Evaluation• Environment– ARM cluster: 4 cubieboard2, 1 head node, 3 worker
nodes; lubuntu for ARM, java-1.7 for ARM– X86_64 cluster: 2 firefly nodes. 1 head node, 1 worker
node; CentOS 6.3, java-1.7• Hadoop 1.2 [1]• Testcases– Loadgen– MDAD (Molecular Dynamics Simulation based on
Hadoop MapReduce [2])
• [1]Apache Hadoop• [2]Chen He, “Molecular Dynamics Simulation based on Hadoop MapReduce” , Master thesis, 2011
![Page 7: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/7.jpg)
Evaluation
• Equivalent Performance– Run program on current device and get
turnaround time– To achive the same turnaround time, how many
new devices we need, or could we this?
• Energy consumption– Kill-a-Watt device to collect ARM cluster energy;– ServerTech PDU for gathering x86_64 cluster
energy consumption
![Page 8: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/8.jpg)
Evaluation
• loadgen
![Page 9: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/9.jpg)
Evaluation
• MDAD
![Page 10: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/10.jpg)
Evaluation
• Loadgen Energy
![Page 11: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/11.jpg)
Evaluation
• MDAD energy
![Page 12: Hadoop mapreduce performance study on arm cluster](https://reader035.vdocuments.us/reader035/viewer/2022062220/556ae29cd8b42a86218b4638/html5/thumbnails/12.jpg)
Conclusion
• We build a ARM cluster which is composed of 4 cubieboard2 cards.
• We setup Hadoop cluster on the ARM cluster• We compared the performance and energy
consumption between two clusters• Based on our current data, we conclude that
ARM cluster is not an alternative choice to replace X86_64.