Data Accelerator (Data Accelerator Goose File System, GooseFS) is a highly reliable, highly available, and elastic Data acceleration service introduced by Tencent Cloud. Leveraging the cost advantage of Cloud Object Storage (COS) as the Storage foundation for the Data lake, it provides a unified Data lake entry for computing applications in the Data lake ecosystem, accelerating performance in massive Data analysis, machine learning, and AI-related business access to Storage. It adopts a distributed cluster architecture with features such as elasticity, high reliability, and high availability, offering a unified namespace and access protocol for upper-layer computing applications, making it convenient for users to manage and transfer Data across different Storage systems.
Product Features
The Data Accelerator has two subproducts: GooseFS and GooseFSx.
Data Accelerator Goose FileSystem
GooseFS aims to provide an all-in-one caching solution, with inherent advantages in leveraging data locality and high-speed cache, unifying storage access semantics, etc. GooseFS plays a core role in the Tencent Cloud Data Lake Ecosystem, "connecting computing above and storage below," as shown below.
GooseFS provides the following features:
1. Cache acceleration and data Locality: GooseFS can be hybridly deployed with compute nodes to improve data Locality, leverage high-speed caching features to solve storage performance issues, and improve the bandwidth for writing to COS.
2. Integrate storage semantics: GooseFS supports protocols such as HDFS and POSIX, and is compatible with multiple cloud storage services including Tencent Cloud COS and Cloud HDFS. It is suitable for various ecosystems and use cases.
3. Unified Tencent Cloud-related ecosystem services: including logging, authentication, and monitoring, unified with COS operations.
4. Provide Namespace management capability: For business data stored in Tencent Cloud COS, Cloud HDFS, and other services, provide different read-write caching strategies and lifecycle management capabilities.
5. Perceive Table metadata functionality: For data Tables in big data scenarios, provide GooseFS Catalog to perceive metadata Tables, and offer Table-level Cache preheating.