bitsdoor.com

UPS is deployed on-premises as either a virtual machine, a container or a hardware appliance, enabling your applications to access both private and public cloud storages, such as AWS, Azure, and etc. By supporting protocols like NFS, S3, and SMB, it delivers highly optimized data transfer along with a rich set of features. Key benefits include automatic data duplication to increase reliability, low-latency storage performance, built-in encryption and data compression, automatic data tiering to help reduce storage costs and etc

FAQ

Q: What is the USP?

A: The Unified Storage Platform(USP) is a storage device management platform that can manage a variety of storage devices (such as NAS, distributed object storage, tape libraries, and blue-ray storage). By abstracting the differences between these devices, it provides file and object storage services to upper-layer applications.

Q: What is the relationship between USP, upper-layer applications, and storage devices?

A: USP sits between the upper-layer applications and the underlying storage devices. From a logical perspective, USP acts as a large storage device. All read/write operations from the applications interact only with USP, which then performs the actual data read/write operations on the underlying storage devices.

Q: How does UPS differ from AWS Storage Gateway and other similar products?

A: While UPS offers functionality comparable to AWS Storage Gateway, it stands out by simultaneously managing storage systems from multiple vendors—including AWS, Azure, NAS devices, and privately deployed object storages—through a unified interface.

Q: What benefits does USP bring to customers?

A: USP can be applied in multiple scenarios. For example, it ensures high availability by writing data across multiple storage systems; it automatically classifies data into hot and cold tiers to reduce storage costs; it improves management efficiency through deep integration with customers’ storage operation platforms; and it enables protocol conversion between NAS and S3 to avoid modifications to front-end applications. For detailed information about USP’s use cases, please refer to the Solution section on our website or contact us.

Q: Is USP a software or hardware product?

A: USP is a software product that can be deployed in various environments, including physical machines, virtual machines, and containers.

Q: Does USP require special interfaces or permissions when managing underlying storage devices?

A: No. USP only requires basic read, write, and delete permissions for the storage devices it manages. USP treats each managed storage device simply as a space for data storage.

Q: How does the upper-layer application interact with USP?

A: From the perspective of upper-layer applications, USP appears as a storage device. Therefore, when using and integrating with USP, applications can treat it exactly like a storage device.

Q: How many files can USP manage?

A: Theoretically, USP has no limit on the number of files it can manage. Based on practical application experience, it is recommended that a single cluster deployment should not include more than 64 nodes, and the total number of managed files should not exceed 24 billion.

Q: How does USP achieve unified management of heterogeneous storage resources?

A: This is a complex question, but in simple terms, it involves the following aspects:

USP requires basic read, write, and delete permissions on the storage devices it manages, allowing it to respond to and fulfill various read/write operations from upper-layer applications.

USP centrally manages data metadata, so logically, USP appears as a single storage device.

Q: What happens if one of the storage devices managed by USP becomes full?

A: USP can monitor the space usage of each storage device and is equipped with load balancing capabilities to ensure even utilization across all devices. Therefore, it's unlikely that one storage device will become nearly full while others remain underutilized.

Q: What impact would a USP failure have on the system?

A: USP is typically deployed in a clustered environment in production environments. A failure of a single node will not affect the normal operation of the entire system. In extreme cases where the entire USP cluster becomes abnormal, services may be interrupted, but the data remains safely stored in the storage devices managed by USP. USP itself does not cause data loss.

Q: What are the hardware resource requirements for USP?

A: USP can be deployed in various environments including physical servers, virtual machines, and containers. Overall, its hardware resource requirements are relatively low. CPU and memory are only used to run software services. USP requires SSDs to store metadata, and the capacity required depends on the number of files being managed. For example, if managing 1.2 billion files, we recommend using 960GB of SSD storage. You can also contact us for a comprehensive planning recommendation.