Help & Documentation>Elastic MapReduce>Product Introduction>Overview

Overview

Last updated: 2023-12-21 15:31:52

On this page:

Elastic MapReduce (EMR) is a secure, cost-effective, and highly reliable open-source big data platform, founded on cloud-native technology and the expansive Hadoop open-source ecosystem. It offers easily deployable and manageable open-source big data components such as Hive, Spark, HBase, Flink, StarRocks, Iceberg, Alluxio, and more, assisting clients in efficiently constructing cloud-based, enterprise-level data lake architectures. EMR supports deployment based on two types of resources: CVM and TKE.

Supported formats

Tencent Cloud's EMR offers two deployment and operation methods based on Cloud Virtual Machine (CVM) and Tencent Kubernetes Engine (TKE):

Modalities
Description
EMR on CVM
EMR is responsible for installing and deploying open-source big data components on CVM, and initiating the corresponding services. You can carry out operations and maintenance tasks for the cluster CVM and services through the EMR console.
EMR on TKE
If you have prepared a standard TKE cluster or a Serverless cluster, EMR will install and deploy open-source big data components based on TKE resources, realizing the containerized operation of the open-source big data platform, thereby reducing your operational focus on underlying resources.

Associated Cloud Products

Feature Name
Note
Under the product form of EMR on CVM, CVM instances are utilized for EMR cluster nodes.
CVM nodes and POD nodes can be paired with cloud block storage of varying specifications for data storage as required.
Under the product configuration of EMR on TKE, one can opt for PODs within the TKE cluster to serve as EMR cluster nodes.
When components such as Hive, Hue, and Ranger are deployed in the EMR cluster, the cluster will concurrently purchase a Cloud Database MySQL for storing component metadata.
In scenarios of storage-computation separation, EMR can be utilized to read and write data in object storage.
In storage-computation separation scenarios, EMR can be employed to read and write data in Cloud HDFS.
Under the product form of EMR on TKE, certain services can be configured with a load balancer to facilitate external access.
Grant service role permissions to the EMR product or authorize operations for collaborators/sub-users through Access Management.
Metric and event monitoring data from the EMR cluster are reported to TCOP. Through TCOP, cluster monitoring data can be retrieved and alarm notification strategies can be configured for key observability metrics.

中国站 - English