Scaling Up Master Node Specifications

MRS provides Manager for managing clusters and services in the clusters, such as NameNodes of HDFS, ResourceManagers of Yarn, and Manager management services of MRS, are deployed on the Master node of the clusters.

With the rollout of new services, a cluster scale increases continuously, and Master nodes bear more and more loads. Enterprise users are faced with the problem that CPU loads are too high and memory usage exceeds the threshold. Generally, in an on-premises big data cluster, you need to migrate data and purchase hardware with advanced configurations to scale up the Master node specifications. MRS leverages the advantages of cloud services to enable you to scale up Master node specifications in one click. During the scale-up, the active/standby HA mode of the Master nodes ensures that existing services are not interrupted.

For details about how to scale up the Master node specifications, see Scaling Up Master Node Specifications.