Distributed Storage APIs using HBASE and HDFS for Encrypted Personal Health Records

Abstract

This work proposed a distributed storage (DSEPHR) system for storing encrypted personal health record data (PHR) by identifying properties of the encrypted PHR data and using the properties to create an index for the encrypted PHR data. The encrypted PHR data contain health related information of an individual. Both HDFS and HBase are used as a fundamental storage framework. The DSEPHR provides a set of APIs for storing and retrieving the encrypted PHR data. The DSEPHR eliminates a high memory consumption issue on the cloud storage caused by storing a lot of small files. The experimental results showed that the DSEPHR still preserve the scalable feature while consuming less memory. Furthermore, DSEPHR can deliver similar upload / download time performance when the mixture of file size and the ratio of read and write activities change.