• DHANAPRIYA L
• Dr. V. UMARANI

Big-data systems allow storage and examination of huge amounts of data, and are fuelling the data revolution that is impacting roughly all walks of human undertaking today. The establishment of any big-data system is an extensive, circulated, data storage system. To increase in the huge amount of data to be read and transferred during recovery for an erasure-coded system results in two major problems: high degraded read latency and longer reconstruction time. To overcome these problems, this paper presents an enhanced method of Erasure-Coded Data Archival System for Hadoop Clusters using Enhanced Greedy Block Failure Recovery Algorithm (EGBFRA), which combines Spark framework and load balancing strategy to solve the optimization problem during Big data processing. The proposed system implemented using Spark 1.6 with Hadoop 2.6 platform under goes MapReduce framework. According to the EGBFRA strategy it can work with the large file database for finding the mining data block failures. The results shown that the performance of the new scheme is effective compared with other aHDFS algorithms. As the Experimental results show, Enhanced Greedy Block Failure Recovery Algorithm clearly outperforms HDFS-EC and aHDFS.

Hadoop distributed file system, Spark, erasure codes, Greedy search

"DESIGN AND DEVELOPMENT OF ERASURE-CODED DATA ARCHIVAL USING ENHANCED GREEDY BLOCK FAILURE RECOVERY ALGORITHM ON SPARK RDD FRAMEWORK", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 6, page no.168-173, June 2019, Available :http://www.jetir.org/papers/JETIR1906L24.pdf

"DESIGN AND DEVELOPMENT OF ERASURE-CODED DATA ARCHIVAL USING ENHANCED GREEDY BLOCK FAILURE RECOVERY ALGORITHM ON SPARK RDD FRAMEWORK", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 6, page no. pp168-173, June 2019, Available at : http://www.jetir.org/papers/JETIR1906L24.pdf