Combining cloud computing and community open-source technologies such as Hadoop, Hive, Spark, HBase, Presto, and Storm, Tencent Cloud Elastic MapReduce (EMR) provides secure and cost-effective cloud-based Hadoop services featuring high reliability and elastic scalability. Using EMR, you can create a secure and reliable Hadoop cluster in just minutes to analyze petabytes of data stored on the data nodes in the cluster or in Cloud Object Storage (COS).
Elastic MapReduce
An elastic and open-source cloud-based Hadoop service.
Overview
Benefits
Open-Source
Delivers high-performance, enterprise-grade and reliable open-source big data ecosystem (Hive, Spark, Presto, HBase, Flink, Iceberg, Alluxio, etc.) with on-demand component orchestration.
Efficient O&M
Cloud-native unified observability and critical event snapshot rollback enable streamlined troubleshooting and improved O&M efficiency.
Elastic Resource Scaling
Auto-scales cluster compute resources seamlessly based on schedule or workload in minutes to adapt to dynamic business scenarios.
Expense Optimization
Pay-as-you-go resource allocation, simplified deployment/maintenance, and support for compute-storage decoupling.
Features
Quick Deployment
Launch a dedicated cluster in just 3 steps via the EMR console. EMR offers a comprehensive suite of open-source components (Hive, Spark, HBase, Presto, Flink, etc.) that can be flexibly mixed and matched during cluster creation to meet custom business needs.
Cluster Creation and Scaling in Minutes
Deploy a secure, stable cloud-based Hadoop cluster in just minutes via the console. Or you can seamlessly scale existing EMR clusters within minutes to accommodate growing data volumes and business demands.
Compute-Storage Decoupling
Decouple storage and compute nodes in cloud-hosted Hadoop clusters. Flexibly scale compute nodes on demand to reduce hardware costs. Or leverage Tencent Cloud COS (Cloud Object Storage) for storage-computation decoupling, enabling independent scaling of storage and compute resources to better adapt to fluctuating business needs.
OPS Support
EMR features a comprehensive monitoring and OPS system that instantly detects exceptions in core components (Spark, Hive, Presto, etc.) and running tasks, safeguarding the stable operation of clusters
Scenarios
Cloud Migration
Cloud-Native Data Lake
Offline Data Analytics
Interactive Business Query
Streaming Data Processing
Self-managed components built on IDC or open-source Hadoop distributions often suffer from complex technology stacks, outdated component versions, high operation and maintenance overhead, and limited technical support. EMR enables seamless migration with a comprehensive suite of migration tools, allowing rapid deployment of cutting-edge, stable, high-performance, and cost-effective cloud-native big data platforms.
As enterprises continue to unlock the value of full-scale data, traditional architectures are unable to meet the demands for cost-effective, unified data storage and management, nor can they flexibly support data analytics tasks across diverse scenarios. Cloud-native data lakes built on EMR effectively address these challenges.
Offline data warehouses centered on Hadoop can leverage tools like Hue to access mainstream compute frameworks (including Hive, Spark, and Presto), enabling rapid derivation of actionable data insights.
In sectors such as Internet finance, gaming, and O2O, there is a pressing need for efficient analysis of structured and semi-structured data—such as user behavior, system logs, and orders. With its rich suite of compute components, minute-level cluster provisioning, and horizontal scaling capabilities, EMR supports real-time interactive business queries and improves business responsiveness.
After pushing real-time data generated from business servers to messaging middleware via APIs and SDKs in applications or tools, you can select the appropriate streaming data processing engine within EMR to analyze the data, enabling real-time data computation and decision-making.
Pricing
Tencent Cloud Elastic MapReduce (EMR) offers flexible deployment and billing options. Billing is granular to the node level—you can select nodes of varying specifications to form clusters, and scale dynamically to match business fluctuations. For details, refer to Billing Overview.