Draft:HBlock

"An innovative lightweight storage cluster controller launched by esurfing cloud."

Description
Developed by esurfing cloud, HBlock is a Storage Resource Reutilization System (SRRS) that improves resource utilization, optimizes resource costs, and assists enterprise users in achieving green transformation. It is a lightweight software-defined storage cluster controller that integrates scattered disks in each server into a high-performance storage resource pool. It provides distributed block storage services through the standard Internet Small Computer System Interface (iSCSI) protocol, and ensures low latency, high availability, and elastic scalability through a distributed multi-controller architecture; it has strong compatibility and supports heterogeneous hardwares to make full use of all storage resources.

Product Advantages
Heterogeneous Hardware

HBlock allows each node in the cluster to have different configurations, such as different architectures (x86 and ARM), and different configurations for CPUs, memories, and hard drives, to improve hardware utilization.

Hybrid Deployment

HBlock is a user-mode process-level software-defined storage cluster controller that does not rely on or modify the operating system environment, monopolize the hard drive, or interfere with the execution of any other processes.

Security and Reliability

HBlock supports multi-copy and the Erasure Code (EC) data redundancy protection mechanism, ensure data persistence in the event of a single-point fault or a sub-healthy state.

Low Read-Write Latency

HBlock adopts distributed multi-controller architecture to improve read-write performance with ultra-low latency. The latency of randomly writing 4KiB blocks can be as short as several sub-milliseconds.

Features
Out-of-the-box

The HBlock installation zip package is highly optimized and only about 170 MB. It can be installed on the Linux operating system (OS) with only three commands, and it takes less than 3 minutes from unzipping the installation package to cluster initialization. HBlock can be installed on physical servers, bare metal servers, and virtual machines (VMs).

Elastic Scalability

HBlock supports scale-out from three servers to thousands of servers without interrupting business, and vice versa. After the expansion, there is no need to relocate a large amount of data, and the system can automatically achieve load balance.

Business Continuity

HBlock achieves seamless session transfer with a failover speed reaching the second-level,  comparable to traditional high-end storage. All services in the cluster are deployed in a redundant mode, ensuring that failures in a single server, single link, or individual disk do not affect availability, achieving 24/7 business continuity of enterprise applications.

Private Cloud Storage
Scenario

HBlock can be connected to various local storage resources on the server efficiently to build a private cloud storage resource pool, reduce O&M costs of complex private cloud data centers on the infrastructure layer, unlock hardware locks, and eliminate data silos. It builds a flexible and scalable private cloud storage base of high availability with low cost.

Product Advantages

Resource reutilization: HBlock does not monopolize the hard drive, can be deployed on a set of hardware devices already installed with applications, and does not affect the running of existing businesses. By assisting users in uniformly managing devices with low storage utilization within their existing server clusters, HBlock creates a unified storage resource pool, optimizing the utilization of these resources.

Supported containers: Compatible with Kubernetes CSI, provides storage resources for containers, reduces deployment costs, and supports elastic scalability.

Edge Computing
Scenario Description

Edge computing is a distributed computing paradigm that brings computation and data storage closer to the location where the data is being generated and where actions are being taken, rather than relying on a central location that can be thousands of miles away, to improve response time and save bandwidth.

Pain Points

Running resources are scattered on edge nodes with varied hardware specifications and environmental reliability.

The local storage unit of an edge node has limited capacity and low scalability.

Product Advantages

Specifically designed for chaotic environments and can be applied to various uncertain environments such as unstable networks, unstable computing, and unstable disks.

Supports heterogeneous hardware and allows each Linux OS instance in the cluster to have different hardware configurations.

Able to reuse the existing storage resources of the edge devices, easily expand the storage capacity on demand at any time without interrupting the service, and provide persistent data storage.

Secondary Storage
Scenario Description

Secondary storage is used to store noncritical data that is not frequently accessed and can run effectively on economical and lower-performing devices, which are more suitable for long-term storage. The secondary storage can store backup and disaster recovery data, video surveillance data, and archived data.

Pain Points

The data capacity to be stored are large and require additional large-capacity storage devices or cloud storage resources.

Product Advantages

Built using commodity servers or existing servers to improve resource utilization.

Supports heterogeneous hardware and allows each Linux OS instance in the cluster to have different hardware configurations.

HBlock is a user-mode process-level software-defined storage controller that does not rely on, or modify the operating system environment, monopolize the hard drive, or interfere with the execution of any other processes.