The highlights of MRS are as follows:
MRS provides not only the capabilities supported by Hadoop, Spark, Spark SQL, HBase, and Hive, but also the unified SQL interaction interfaces in the entire process, which simplifies big data application development.
MRS is free of O&M and separates computing from storage. The computing cluster can be created as required and released after a job operation is complete.
MRS reduces time spent on commissioning and monitoring clusters and delivers high availability and reliability.
MRS is open source-based and compatible with other services, and provides REST APIs and JDBC interfaces.