Hyper-converged Architecture-hyper converged Infrastructure Small note

Source: Internet
Author: User

A few days ago, a colleague suddenly came to me and asked me what the hyper-converged architecture is and what it does. At that time I was still in a vague state of the concept, so I failed to give a good answer.

So it took a little time to study the popular technology, and to tell my colleagues once again, the effect is good. Summarize it yourself.

To understand this concept, let's look at the current architecture and some of the challenges that exist. The current architecture, mainly through the network, the server and storage connection, so as to achieve storage sharing. And through the storage of redundant functions, to achieve high availability.

This architecture should be the mainstream architecture so far, which I think is a traditional architecture, but it feels a bit premature. The main problems with this type of architecture are:

    • High system complexity: Because of the app--server-network-storage and other aspects
    • Expansion difficulty: As applications, data, server increases, storage CPU, capacity and other needs of the upgrade, storage needs to be upgraded, or scale-out, which will further improve the complexity of the system
    • Network latency and Complexity: Network part delay, switch configuration, compatibility, etc. cannot be ignored

Hyper-converged architectures can address these issues by changing several important components and architectures.

Simply put, hyper-convergence is the integration of the traditional architecture of the network, the storage part of the software and the way the x86 server. And through the virtualization technology of these servers for unified management. The functions in the operating system of the original storage device are implemented by software. This allows the server, network, and storage to be managed in a unified interface. Nutanix, for example, has four x86 servers in a 2U chassis, with SSD (performance) and HDD (capacity), and CVM (the Controller VM, which is the implementation of the stored operating system), and VMware's ESXi as a hypervisor, it implements a hyper-converged architecture. Storage is managed and used as one or more storage pools. Applications, where the virtual machine accesses data through CVM to access data locally or on other servers. And the data is synchronized between the different nodes. can see:

Let's look at how this architecture solves the challenges of traditional architectures.

    • Through hyper-Fusion, the original multi-level is integrated together, greatly reducing the complexity.
    • If you are faced with a lack of storage resources, you can add more resources directly through a scale-out approach. And the extended process is considerably simplified relative to the previous architecture. And the number can be extended to very large, for example, Nutanix currently does not have a hard limit on the number of nodes
    • Because the network and storage are all fused together, the configuration of the network becomes simple and the latency is drastically reduced compared to the previous ones.

Here, the hyper-converged architecture looks like a great idea. But there are also some problems. For example

    1. The storage management in the form of software is not very rich in function at present.
    2. The extension of hyper-converged systems is the overall expansion, storage, computation, and networking, but if the customer wants to add a portion of the resources separately, hyper-integration in this part of the granularity is not enough. For example, traditional storage can be achieved by simply adding a disk to the enclosure.
    3. Hyper-converged implementations are not limited to architectures such as Nutanix, where VMware's Evo:rail and NetApp FAS2552 are also an implementation of another hyper-converged architecture. Both functional and scalable.

It is a circle in which it feels like a cycle of architecture can occur after a period of change. Hyper-Fusion is no exception, at the time of shared storage just rise, the server is an island, storage is not shared, better use. Now that hyper-converged, some of the challenges of shared storage have turned back to server consolidation storage, but this time it solves the original server island problem with virtualization technology. Can be understood as a more flexible, more efficient, scalable and easy-to-upgrade version of the architecture.

Hyper-converged Architecture-hyper converged Infrastructure Small note

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.