ucloud global logo
Managed Hadoop (UHadoop)
UHadoop is a comprehensive big data processing platform based on the Hadoop framework. It provides ready-to-use common big data ecosystem components including Spark, HBase, Presto, and Hive, with optional auxiliary tools such as Hue, Sqoop, Oozie, and Pig. To meet compute-storage separation requirements, UHadoop now supports independently managed HDFS storage clusters that can serve multiple independent compute clusters for data read and write.
Product Advantages
Product Features
Related Documents

Product Advantages

  • Low Cost — Same Price as Self-Built Hadoop on Cloud Hosts
    Low Cost — Same Price as Self-Built Hadoop on Cloud Hosts

    UHadoop cluster pricing is currently the same as self-built Hadoop clusters, with plans to progressively reduce costs in more availability zones.

  • Dedicated Physical Node Options for Lower Cost and Better Performance
    Dedicated Physical Node Options for Lower Cost and Better Performance

    In addition to standard virtual machines, dedicated bare-metal physical nodes without virtualization are available, supporting PB-scale massive storage and high read/write performance scenarios.

  • Shared HDFS Storage for Balanced Read/Write Performance and Cluster Flexibility
    Shared HDFS Storage for Balanced Read/Write Performance and Cluster Flexibility

    Multiple compute clusters can access data from the same HDFS storage cluster. This model reduces resource costs for large-scale clusters while allowing compute cluster services to run more stably.

  • Superior Cluster Stability with UHadoop Agent Monitoring and Auto-Recovery
    Superior Cluster Stability with UHadoop Agent Monitoring and Auto-Recovery

    No need to worry about node disk failures or other low-level issues, or whether cluster services are available. The in-cluster Agent monitors the cluster and automatically recovers from failures in a robust manner.

  • Console-Based Service Component Management for Improved Efficiency
    Console-Based Service Component Management for Improved Efficiency

    The console supports enabling, disabling, and modifying configurations of cluster service components such as Spark, Hive, HBase, and ResourceManager — no need to log in to individual nodes to manage cluster services.

  • Console Visibility into YARN Application Status to Aid Application Development
    Console Visibility into YARN Application Status to Aid Application Development

    The console displays status, logs, and other details for Applications and their sub-tasks submitted within the last 15 days, making it easy to troubleshoot and trace business issues.

  • Flexible Billing to Balance Processing Efficiency and Cost
    Flexible Billing to Balance Processing Efficiency and Cost

    Supports both annual/monthly prepaid billing and pay-as-you-go flexible purchasing.

Product Features

Features
- Supports Hadoop/Spark
- Supports creating data clusters in minutes
- Supports online cluster scale-out and scale-in
- Flexible and configurable node types
- Dedicated disk resources per node
- Supports SAS/SATA disks
- Supports cluster health monitoring and alerting
- Automatic node fault recovery
- Supports API and console operations
- Supports PB-scale data processing
- Supports structured and unstructured data processing
- Supports service components including Hive, Pig, HBase, and Hue
- Supports development environments including Java, .NET, PHP, and C++
- Open root access on nodes for custom software installation
- Exclusive use of cluster resources per user
- Supports pay-as-you-go billing
- Provides data import services