Product Features
High-Performance Computing Framework
Supports high-performance computing frameworks, aiding users in efficiently completing tasks in the data warehouse domain.
High Availability and Flexible Scalability
The big data cluster is supported by the KDP computing resources/storage base, offering features such as high stability, high availability, and high scalability.
Support for Multiple Scenarios and Applications
Gradually implements offline/real-time data warehouses, OLAP analysis platforms, and operational databases through KDP.
Multi-Version Compatibility
KDP has conducted extensive adaptation work to address compatibility issues, supporting perfect compatibility among multiple versions of multiple components.
Visual Monitoring, Operation, and Deployment Capabilities
Provides a visual cluster management, monitoring, and operation interface, enabling cluster management through simple clicks, significantly improving cluster management efficiency.
Features
Function Introduction
The KDP Deployment & Monitoring Platform is designed with the continuation of the master-slave architecture philosophy. The master node is responsible for global coordination, management, and task distribution; while the slave nodes are responsible for executing specific tasks, deploying big data services, monitoring service status, etc. It supports both bare-metal and cloud-native deployment methods.
Functional Modules
Module One: Service Management Module, responsible for the deployment and operation & maintenance (startup, shutdown, upgrading, etc.) of big data services; Module Two: Service Monitoring Module, responsible for monitoring various health indicators and performance indicators of the services; Module Three: Host Management and Service Alerting.
Application Scenarios
High Requirements for Real-Time Data Capability
Business attributes often demand high real-time capability for data. For instance, the securities industry has stringent requirements for real-time trading data. With the KDP-Flink CDC Data Chain Module + Computation Module, data can be accessed and processed within seconds.
High-value Customer Analysis
In sectors like retail, KDP's OLAP capabilities (including Impala, Presto, ClickHouse, etc.) can be utilized to unify metrics and templates, support multi-engine switching, and provide more efficient processing speeds for business needs.
Artificial Intelligence Industry
The artificial intelligence industry focuses on the efficiency of feature training. Traditionally, this involves offline imports using Kafka - HDFS - Spark. However, KDP-Pulsar + OpenML DB enables the transition of feature data from offline to real-time implementation, reducing data transfer time for more timely and accurate training results, thus enhancing efficiency.
Migrate from CDH or HDP
Customers currently using the free versions of CDH or HDP platforms face challenges due to Cloudera's cessation of version maintenance and technical support for these products. KDP offers compatibility with multiple components and versions of domestic Apache Hadoop, fully replacing HDP and CDH.
Operational Excellence
For existing self-built and self-operated big data clusters with substantial accumulated data and a scale of thousands of nodes, leading to high costs, KDP provides a combined solution package of KDP - HDFS StoreLayer + Cloud Storage + Alluxio. This helps decompose the storage layer, utilizing both local and cloud storage to reduce cluster costs.
Product Capabilities
- Containerized standard components support a storage-and-computing-separated architecture, enhancing resource deployment efficiency.
- Licensed for perpetual use after one-time deployment, with flexible scaling up/down capabilities.
- Supports both on-premises and cloud deployments, selectable based on needs.
- Based on a unified batch and stream processing framework, supporting real-time and offline data analysis.
- Supports OLAP and high-performance database engines, catering to diverse application scenarios with excellent scalability.
- Web-based visual interface for easy management of big data clusters.
- Comprehensive encapsulation of operational tasks, shielding complex underlying operations and lowering the threshold for maintenance.
- Fault identification and resolution, with intelligent recognition and one-click solutions for common maintenance issues, such as HDFS metadata recovery, YARN scheduling bottleneck diagnosis and handling, and detection and handling of MR long-tail tasks.
- Authentication, authorization, and authentication mechanisms based on Kerberos + Ranger, providing comprehensive security protection for enterprise users.
- Supports triple redundancy backups, offering highly reliable commercial data storage services.
- Dedicated 1V1 service with 7*24H online support.
- Professional maintenance team ensuring up-to-date component versions and support for new features and components.
Learn more, start your data intelligence journey now