Hyperconverged Infrastructure (HCI) Now Runs on IBM Power Systems

Stefanie ChirasToday’s corporate data centers are using more and more compute and storage resources to meet rapidly increasing operational requirements. Because traditional data center architectures are experiencing great difficulty in meeting these new demands, an alternative technology is swiftly gaining acceptance. Hyperconverged Infrastructure, or HCI, is on what Stefanie Chrias, IBM’s VP Power Systems, calls “a rapid growth trajectory.” And now, for the first time, this new technology that is so swiftly penetrating enterprise data centers is available to run on IBM’s Power Systems platforms.

But what, exactly, is hyperconverged infrastructure?

HCI takes the fundamental elements of the data center, servers, data storage, and networking, and packages them together in a single unified appliance. The entire unit, as well as its component parts, is controlled entirely by sophisticated software under the direction of detailed policies established by IT administrators. Both the compute engine and the storage controller run on the same server platform, and each appliance functions as a node in a cluster.

The constituent parts of the HCI appliance are hidden behind a unified “single pane of glass” software interface. So, there is no need for users or applications to deal directly with the hardware or its particular characteristics. The software can automatically and transparently carry out tasks such as performing data backups, scaling out (simply by adding nodes) to provision additional storage as needed, or swapping out nodes that fail. This approach greatly simplifies the IT management task.

Part of the appeal of HCI is that it was designed to run on inexpensive industry-standard x86-compatible servers and storage devices. But that meant IBM’s RISC-based Power Systems line was shut out of this fast-growing market.

HCI and Power Systems

IBM Power SystemsNow, however, IBM has announced that it is partnering with Nutanix, which 451 Research has named as the leading HCI provider, to market appliances based on the Power Systems line rather than x86 servers. Because of the superior compute and data handling capabilities of the Power architecture, IBM believes this new platform will allow enterprise customers to “run any mission critical workload, at any scale, with world-class virtualization and automation capabilities.” The platform is particularly suited to running high performance database, analytics, machine learning, and artificial intelligence applications.

For IBM, HCI is “a fundamentally different approach to enterprise application needs.” It also represents an important emerging market that IBM didn’t want to be left out of.


Looking for Power Systems education? Take a look at COMMON’s online offerings.

Layers of a Scalable Cloud Architecture

Cloud

The cloud computing ecosystem is huge and consists of several technologies. Many companies rely on these varying cloud infrastructure to deliver their products and services efficiently. This brings up the question, how scalable is your cloud architecture?

Using the right architecture is extremely crucial for your entire cloud’s operation. It is important that organizations understand the specific requirements of their servers, and if they are already using a cloud platform, decide on the type of cloud architecture that would be best for their business logic.

Before choosing a cloud computing architecture, the first thing that’s required is a scalable structure. Cloud computing is scalable when all its components are independent of each other. This independence allows systems to scale at exceptional levels and is usually implemented at the design stage.

Features of a Scalable Cloud Architecture

Typically, cloud computing systems involve different cloud components communicating with each other on a system that functions like a messaging lineup. How these components interact is what determines the scalable nature of your infrastructure. There are two layers that make up a scalable cloud architectures:

1. The Client / Front-end

The client structure is where all users interface with the target platform. This is usually the mobile or web application that manages users, sessions and pages. The client usually makes API calls to the server.

The front-end comprises of single user or a network of users. Note that some front-ends will not look like the regular applications we see everyday. The main thing to remember during the design stage is that this is the layer that communicates with the back-end. Therefore RESTful calls to the back-end is the main purpose during the front-end design stage. Whatever visual design you build into your cloud’s front-end, making API calls to the server is the main focus at this stage.

2. Server / Back-end

Your server comprises of data, caching services and all services that interact directly with your server applications. This interaction is necessary for data delivery.

Your server applications drive your business functions and can include apps like CRM, inventory, accounting, reservation system and much more. Adding new applications is part of scalability, so as you add new applications, the demands of higher traffic and computing loads must be anticipated. Your front-end will not automatically scale to size unless you ensure that your back-end accommodates the new load and traffic.

For best practices in maintaining and protecting client’s data, a cloud computing structure requires a high level of redundancy than is necessary for a system hosted locally. The backup created by this redundancy means that the back-end server can jump in and access backup images for quick restoration of data.

In a highly scalable cloud computing architecture, applications are managed, controlled and served by the back-end. The strength of the back-end is how it manages security protocols, traffic and system files. If the applications on your server are broken down and classified into sub-components of the main server, your cloud infrastructure will deliver limitless efficiency and possibilities, making scalability much easier.

The Move to Cloud Storage is Gathering Steam

Both large enterprises and smaller companies are well into a process of transitioning their data storage from on-premises data centers to the cloud. The advantages of cloud-based IT storage are compelling.

Advantages

First, cloud storage relieves a company of the necessity of up-front capital spending to purchase storage hardware. Cloud storage vendors store the data in their own remote data centers and charge a monthly fee for just the amount of storage a customer actually uses in that billing period.

Another advantage is that good cloud storage providers are experts in data management, data security, backups and disaster recovery. This takes a tremendous load off a customer’s in-house IT staff, which can then devote its attention to issues that are more focused on the company’s core businesses.

Concerns

One of the major concerns many IT managers have about moving to the cloud is the security of their data when it resides in someone else’s facilities. But that concern is being alleviated by two factors.

The first is that due to the nature of their business, cloud storage providers are necessarily extremely competent at protecting their clients’ data, and can usually do a better job of it than all but the largest companies can do on their own.

The second factor that mitigates concern about using cloud storage is that it’s not an all-or-nothing proposition. Many companies have settled on a hybrid solution that keeps their most business-critical data at home in their own data centers, while farming out less sensitive data to a cloud storage provider. Many times a cloud storage vendor actually implements and manages a private cloud on the customer’s premises. A survey by 451 Research indicates that by the end of 2014, 39 percent of enterprises had already moved to a hybrid solution.

Outlook

In 2015 spending by hardware manufacturers such as IBM on equipment destined for use in cloud data centers grew by 22 percent to almost $29 billion. According to IDC, spending on cloud infrastructure is expected to reach $37.5 billion by 2020.

As they become more familiar with the advantages of cloud computing, companies big and small are deciding that the cloud should be an integral part of their data storage infrastructure.

Cloud