An Overview of Hadoop MapReduce, Spark, and Scalable Graph Processing Architecture
| Autor: | Karishma P. Talan, Kartik U. Sharma, Pooja P. Talan, Pratiksha P. Nawade |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Computer science
business.industry Distributed computing Big data 020206 networking & telecommunications 02 engineering and technology Software_PROGRAMMINGTECHNIQUES Stream processing Open source Scalability Data_FILES 0202 electrical engineering electronic engineering information engineering Batch processing Graph (abstract data type) 020201 artificial intelligence & image processing InformationSystems_MISCELLANEOUS Architecture business Drawback |
| Zdroj: | Advances in Intelligent Systems and Computing ISBN: 9789811312793 |
| DOI: | 10.1007/978-981-13-1280-9_3 |
| Popis: | In today’s technology era, Big Data has become a buzzword. Various frameworks are available in order to process this Big Data. Both Hadoop and Spark are open source framework to process Big Data. Hadoop provides batch processing while Spark supports both batch as well as stream processing, i.e., it is a hybrid processing framework. Both frameworks have their own advantages and drawback. The contribution of this paper is to provide a comparative analysis of Hadoop MapReduce and Apache Spark. In this paper, we also propose a scalable graph processing architecture that could be used to overcome traditional limitations of Hadoop framework. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|