TY  - JOUR
T1  - Big Data Optimization for Social Networking Tweet
AU - Senthikumar, M. AU - Ilango, P. 
JO  - International Journal of Soft Computing
VL  - 11
IS  - 5
SP  - 305
EP  - 311
PY  - 2016
DA  - 2001/08/19
SN  - 1816-9503
DO  - ijscomp.2016.305.311
UR  - https://makhillpublications.co/view-article.php?doi=ijscomp.2016.305.311
KW  - HDFS
KW  -hadoop
KW  -well-known platform
KW  -response time
KW  -trillions
AB  - Over trillions of active users are there in the social media network worldwide, generating insurmountable data and of dynamic structure. That is why social media is indeed mountains of big data to be explored and performance is a great issue to be improved. Apache Hadoop (a cloud service) is a well-known platform for its scalability, fault-tolerance and capability of processing big data. Hadoop MapReduce gives users full control on how input datasets are processed. Hence, Hadoop is the heart of big data analytics. However, there are several issues challenging Hadoop performance. Hadoop has a large set of configuration parameters which have an impact on performance and successful completion of loads of Hadoop jobs. This study aimed at setting up and tuning Hadoop clusters on performance metrics like workload balance, throughput, network bandwidth and response time. This is achieved by the iterative process of tuning the Hadoop parameters.
ER  -