What Is Hypopyon?

Hyper Converged Infrastructure (HCI) refers to not only resources and technologies such as computing, network, storage, and server virtualization in the same set of unit devices, but also backup software, snapshot technology, and duplication. Data deletion, online data compression, and other elements, and multiple sets of unit devices can be aggregated through the network to achieve modular seamless scale-out and form a unified resource pool. Hyper-convergence (or even hyper-convergence) is easy to implement locally: compute, network, and storage are all integrated into one device, and through it are pre-configured by the vendor and almost ready to use. HCI is the ultimate technology approach to the Software Defined Data Center (SDDC). HCI is a large-scale infrastructure model similar to Google and Facebook's backend, which can bring optimal efficiency, flexibility, scale, cost and data protection to data centers. ^[1]

Hyper-Converged Infrastructure is an information infrastructure that integrates virtual computing resources and storage devices. In such an architectural environment, not only resources and technologies such as computing, networking, storage, and server virtualization are provided in the same set of unit devices, but multiple sets of unit devices can be aggregated through the network to achieve modular seamless horizontal expansion. Out) to form a unified resource pool. ^[2]

Hyper-converged systems integrate storage and computing functions into a single node (or cluster of nodes, each node provides computing and storage functions). Hyper-converged systems have the following common core components: ^[3]

The hyper-converged infrastructure inherits some of the characteristics of the converged architecture. It is also based on a common hardware server. Multiple servers are formed into a cluster with a unified storage pool across nodes to achieve the performance and capacity scalability required by the entire virtualized environment And data availability, you can expand the computing performance and storage space of the entire cluster by increasing the number of nodes in the cluster, and provide high availability and data protection capabilities through the replication and backup of data between each node in the cluster. In order to flexibly allocate resources, the hyper-converged architecture also uses a virtual machine (VM) as the core and a software-defined method to plan and use the underlying hardware resources, and then delivers the required resources to the end users. ^[4]

Server virtualization

Server virtualization is a core component in the entire hyper-converged architecture. Virtualized programs based on bare metal architecture run directly on the server to achieve

1.Minimalist architecture

The hyper-converged architecture directly distributes storage to each PC server, and deploys fast flash disks and large-capacity traditional mechanical disks on the server to meet the system's high IO requirements and large-capacity storage requirements. Therefore, the hyper-converged architecture enables high-speed access to local data without requiring access across networks. The hyper-converged architecture also includes elements such as backup software, snapshot technology, deduplication, and online data compression. Multiple sets of unit devices can be aggregated through the network to achieve modular and seamless scale-out to form a unified resource pool. , Its expansion mode can be changed to add nodes horizontally. With this standardized module, it is undoubtedly very convenient to build a data center. This not only greatly facilitates the construction and management of customers, but also enhances the flexibility of the system, while making deployment and operation and maintenance easier. ^[6]

The rapid development of hyper-converged infrastructure allows administrators to choose among more virtualized storage solutions.

Although there are multiple software-defined storage and storage virtualization solutions on the market today, the nature of storage has not changed. The size and number of virtual machines will not decrease. For enterprises that now have hyper-converged infrastructure, deploying and managing dedicated hyper-converged infrastructure and networks is a valuable experience. Although this approach usually requires a large investment in hardware and personnel, reliability and performance are also commendable. Today, hyperconverged infrastructure products can also provide many related functions, such as data mirroring, deduplication, encryption and disaster recovery.

The emergence of converged and hyper-converged infrastructure has brought storage confusion for enterprises: how to choose the ideal storage solution for the existing virtualization environment and ensure that it can adapt to future development trends? To make the right decision, administrators need to understand the difference between converged and hyper-converged infrastructure and what role traditional storage hardware will continue to play.

The hyper-converged infrastructure uses computing, network, and storage resources as basic components, and is selected and predefined according to system requirements. This method is based on existing hardware modules, so it can achieve rapid expansion, and if different modules are compatible with each other, the expansion process will be very simple. Pre-configured products such as hyper-converged infrastructure will generate a lot of overhead in the early stage, because enterprises not only need to pay for technical costs, but also need to ask vendors to complete related integration work. Over time, the total cost of ownership of a hyper-converged infrastructure can be controlled within a reasonable range. Users have some flexibility in changing and upgrading basic components, but only to the extent allowed by hyper-converged infrastructure vendors. In a hyper-converged infrastructure, storage controllers are still hardware devices and are controlled by "traditional" storage area network (SAN) or network-attached storage (NAS) solutions. This approach is more similar to packaging traditional data center components in a dedicated container.

"Super" converged infrastructure means taking advantage of the existing advantages of converged infrastructure and adding additional software layers, encapsulating computing, network, and storage resources in a single container instead of being provided as a basic component. With the help of Nutanix or EVO: RAIL and other related products, the hyper-converged infrastructure can eliminate the use of storage controllers in storage hardware and add this functionality to software services at the hypervisor layer. Converting storage controllers into software services provides greater flexibility in terms of functionality. This "decentralized" hyper-converged infrastructure reduces the complexity of dedicated hyper-converged infrastructure and networks. While reducing the local storage overhead, it provides many advanced features such as replication and deduplication. With the ability to solve many storage issues in the enterprise, converged and hyper-converged infrastructure products have gained a lot of attention. If a hyper-converged infrastructure attempts to use standard servers in a storage platform, it will encounter problems with hardware endurance and performance. Therefore, these problems can be solved by applying software-defined storage features in the hyper-converged infrastructure, because the main purpose of the hyper-converged infrastructure is to handle the load in the virtual environment.