Hyperconverged infrastructure offers next step in on-premises and cloud-integrated virtualization, as manufacturing continues to advance.
By Alan Conboy, Office of the CTO, Scale Computing
Building smart factories with intelligent manufacturing and business processes have become a top priority from the data center to the C-suite, due to the potentially huge gains that can be enjoyed across areas such as productivity, innovation and product quality, as well as energy efficiency, and even manufacturing employment opportunities.
For the industry’s IT managers, achieving this really comes down to the ability to digitally transform their data center. And, for the majority that are moving towards this goal, hyperconvergence delivers a framework that can provide undeniable and unparalleled benefits, particularly around the industry’s burgeoning use of IoT and edge computing.
IoT (Internet of Things) or commonly referred to as IIoT (industrial IoT) in manufacturing presents the industry with the chance to reap huge gains across industrial processes, supply chain optimization, and so much more. Edge computing and IoT allows manufacturing organizations to decentralize the workload, and to collect and process data at the edge or nearest to where the work is actually taking place, which can overcome the “last mile” latency issues; in addition to minimizing complexity and allowing for easier collection and initial analyzing of data in real time.
Edge computing can also be utilized to offload processing work close to end users, serving as a liaison between the IoT edge devices and larger enterprises hosting the high-end compute resources, for a more in-depth processing and analytics. Unfortunately, numerous manufacturing organizations have encountered hurdles that are seemingly impossible to overcome as they have worked to deploy, manage and enjoy the benefits of IoT and edge computing. And, that’s exactly where hyperconvergence can make all of the difference.
So what’s holding everyone back? Why hasn’t the entire industry already embraced it? The answer is simple. The regular misuse and misunderstanding of the term hyperconvergence has led to uncertainty and continues to act as a barrier for those that could otherwise benefit enormously from an IT, business agility and profitability standpoint. Let’s try to clear up that misunderstanding here.
The 3-2-1 System Doomsday Architecture
Before the emergence of hyperconvergence entered, there was and remains the inverted pyramid of doom, which refers to a 3-2-1 model of system architecture. While it may get the job done in a few areas, it is actually the exact opposite of what a business wants or needs today.
The 3-2-1 model consists of virtualization servers or virtual machines (VMs) running three or more clustered host servers, connected by two network switches, backed by a single storage device – most commonly, a storage area network (SAN). The issue here is that the virtualization host depends completely on the network, which in turn depends totally on the single SAN. In other words, everything rests upon a single point of failure – the SAN.
The Emergence of Hyperconvergence
When hyperconverged infrastructure (HCI) was first introduced, it meant a converged infrastructure solution that natively included the hypervisor for virtualization. The “hyper” wasn’t just hype as it is today. This is an important distinction, as it has particular implications for how architecture can be implemented for greater storage simplicity and efficiency.
Who can provide a native hypervisor? Anyone really. Hypervisors have become a market commodity with almost no feature difference. With free, open source hypervisors like KVM, anyone can create a hypervisor that is unique and specialized to the hardware they provide in their hyperconverged appliances. Many vendors still choose to stay with converged infrastructure models, perhaps depending on the market dominance of VMware―even with many consumers fleeing its high licensing pricing.
Budget benefits is only one of the many advantages of hyperconverged infrastructure. By utilizing a native hypervisor, the storage can be architected and embedded directly with the hypervisor, eradicating inefficient storage protocols, file systems, and VSAs. The most efficient data paths enable direct access between the VM and the storage; this has only been achieved when the hypervisor vendor and the storage vendor are the same. When the vendor owns the pieces, it can design the hypervisor and storage to directly interact, resulting in a huge increase in efficiency and performance.
In addition to storage efficiency, having the hypervisor included natively in the solution removes the need for another vendor which increases management efficiency. A single vendor that provides the servers, storage, and hypervisor makes the overall solution much easier to support, update, patch, and manage without the traditional compatibility issues and vendor finger-pointing. Ease of management represents a huge savings in both time and training from the IT budget.
For example, one company that turned to hyperconvergence was KIB Electronics, a leading manufacturing company of high-quality electronics since 1975. Having already developed an ad hoc network with various equipment purchased from eBay, due in large part to budgetary requirements, KIB Electronics decided to upgrade its IT environment after a large storm knocked out its production servers, resulting in manufacturing and delivery delays. This incident led the company recognize that it needed to have a more reliable data center infrastructure to support its 24×7 uptime requirement, in order to meet its customer-centric goals. After exploring potential solutions that would help it avoid future downtime, it was decided that hyperconvergence would be the ideal solution.
Manufacturing and the Cloud
Most manufacturing organizations have been leveraging the cloud for some time, whether from an on-premises, remote or public cloud platform, or usually a combination of more than one (i.e. hybrid-cloud).
Hyperconverged infrastructure, as a fully functional virtualization platform, can virtually always be implemented together with other infrastructure solutions, as well as integrated with cloud computing. For instance, with nested virtualization in cloud platforms, a hyperconverged infrastructure solution can be extended into the cloud for a unified management experience.
Not only does a hyperconverged infrastructure work alongside and integrated with cloud computing but it offers most of the benefits of cloud computing in terms of simplicity and ease-of-management on premises. In fact, for most manufacturing organizations, a hyperconverged infrastructure may be the private cloud solution that is best suited to their unique environment.
A hyperconverged infrastructure, like cloud computing, is so easy to manage that it allows IT administrators to focus on apps and workloads rather than managing infrastructure and fighting fires all day as is common in 3-2-1 deployments. A hyperconverged infrastructure is not only fast and simple to implement, but it can be scaled out quickly when required. A hyperconverged infrastructure should absolutely be considered along with cloud computing for data center modernization.
Hyperconverged – A Data Center in a Box
After the initial cabling and minimal networking configuration, hyperconverged infrastructure has all of the features and functionality of the traditional 3-2-1 virtualization architecture (except that single point of failure) – which is why it is aptly referred to as a “data center in a box.”
Up and Running in Under an Hour
Time is money and with a hyperconverged infrastructure, time is always saved. Hyperconverged systems can be deployed more quickly than other virtualization solutions because of the appliance-based architecture. Racking and networking are usually the most time-consuming factors in implementation. Deployment times vary by vendor, especially if there is a third party hypervisor to install and VSAs to configure but with a native hypervisor pre-loaded, an entire cluster of appliances can be up and running in under an hour.
IT management in manufacturing organizations typically spend a great deal more time than should be necessary on tedious and repetitive infrastructure management activities and fighting fires. Hyperconverged infrastructure solutions eliminate much of this pain, so IT management can focus on activities that more directly tie to the bottom line. A hyperconverged infrastructure solution can be managed from a single management interface, eliminating the multiple management consoles and interfaces found in 3-2-1 architecture. This is not always the case for hyperconverged infrastructure solutions using third-party hypervisors which typically end up using two interfaces however. For hyperconverged infrastructures with a native hypervisor included, this single interface approach dramatically reduces management time and effort and streamlines management tasks for the administrator.
Native Backup and Disaster Recovery (DR)
IT management should seek hyperconverged solutions where backup and DR are included natively at no extra cost. This enables IT management to eliminate yet another vendor from the IT environment (too many vendors add complication and cost, and open the door to configuration issues). Unlike third-party solutions, native solutions are typically embedded in the storage layer and allow innate awareness of block changes for cleaner backup, replication and recovery options.
High Availability (HA)
While a hyperconverged infrastructure can at times be deployed as a single appliance for individual use cases, it is typically deployed as a cluster of appliances for HA. This way, not only can an appliance absorb the loss of a disk drive, but the cluster can absorb the loss of an entire appliance. Clustering also enables the hyperconverged infrastructure system to scale seamlessly by adding more appliances to the cluster. Some hyperconverged infrastructure solutions require clustering appliances of the same model and configuration, while others allow clustering of heterogeneous appliances.
Simplified Software and Hardware Updates
Performing continuous system software and firmware updates is typically a mind-numbing task, but hyperconverged infrastructures typically make this an easy process. By owning the entire virtualization, server and storage stack, and operating in a HA cluster, updates can be performed automatically across the entire cluster. All software layers (hardware firmware, hypervisor, storage and management) can be upgraded in unison as a single, fully tested system to eliminate component compatibility concerns. VMs can be automatically moved from appliance-to-appliance in the cluster as updates are made to maintain system operations.
Lower Total Cost of Ownership (TCO)
While some may say that a hyperconverged infrastructure may not always be the lowest cost solution in terms of the initial Capex investment. However, more often than not – it is. This is due to its ease of scalability which enables organizations to avoid excessive over-provisioning in the initial investment. Buying only what you need, when you need it, usually leads to significant savings. In addition to CapEx savings, a hyperconverged infrastructure provides considerable OpEx savings over time. This is due to the significant reduction in the costs of management and maintenance. Simplifying an IT environment with a hyperconverged infrastructure can save over 50 percent in the total cost of ownership over 3-2-1 solutions.
Eliminates the Excess Cost and Complexity of 3-2-1 Architectures
Hyperconverged Infrastructure is designed as a replacement for 3-2-1 architecture to eliminate excess cost and complexity. Consequently, it can benefit any size organization in virtually any vertical that requires a robust virtualization environment. However, the extreme simplicity of a hyperconverged infrastructure makes it most beneficial in use cases where dedicated IT staff is limited, as is the case in many manufacturing organizations. Small and medium business (SMB) and distributed enterprises with many remote offices or branch offices (ROBO) commonly have staffing challenges that make a hyperconverged infrastructure the ideal choice.
For a small to mid-sized manufacturing organization, the entire IT staff may be as small as only one full-time or even part-time IT administrator. The complexity of a 3-2-1 architecture can be very challenging, even with the extremely expensive levels of training and certification that accompany it. The simplicity of a hyperconverged infrastructure, on the other hand, can allow it to be managed easily by a junior administrator or allow a more senior administrator to spend less time managing the infrastructure and more time delivering better applications and services, and improving the business.
In a distributed manufacturing organization, remote offices and branch offices seldom have dedicated IT staff. These remote locations often require frequent visits from IT staff which can result in high travel costs and lower productivity. The simplicity of a hyperconverged infrastructure includes multiple redundancies for HA, failure handling and self-healing. A remote site experiencing a failure would not cause an outage nor require immediate replacement, reducing IT staff visits. Greater uptime and accessible remote monitoring and management lead to lower travel costs of IT staff to these locations and significantly lower operating costs―not to mention the increase in productivity.
Hyperconverged Infrastructure – Revolutionizing Manufacturing IT Operations – Enabling Smart Factories
Hyperconverged infrastructure is a revolutionary method for manufacturing organizations to think about IT infrastructure that reduces IT investments in terms of both budget and manpower. While it may be hard to determine whether a solution is truly hyperconverged, or some other pretender, it is worth investigating hyperconverged infrastructure solutions to learn if/how your organization can gain the maximum benefit of modern IT infrastructure.
Not sure where to start? Here are a handful of straightforward questions to ask potential hyperconverged solution providers:
- Does the solution provide a native hypervisor or does it require an additional purchase of hypervisor licensing and support?
- Does the solution offer hypervisor-embedded storage or does it use virtual storage appliances (VSAs)?
- Can the solution combine and scale with heterogeneous appliance models and configurations?
- Does the solution offer native backup and DR capabilities?
- Does the solution integrate with cloud computing and IoT? How?
Hyperconverged infrastructure is the next logical step in on-premises and cloud-integrated virtualization infrastructure, as the manufacturing industry and its associated compute requirements continue to advance. Settling for more traditional virtualization solutions like a 3-2-1 architecture will likely end up costing organizations far more in capital, manpower and headaches than switching over to the simplicity and savings of a hyperconverged infrastructure solution.
About the Author: Alan Conboy, Office of the CTO, Scale Computing (www.scalecomputing.com)
Alan Conboy is the Office of the CTO at Scale Computing since 2009. With more than 20 years of experience, Conboy is an industry veteran and technology evangelist specializing in designing, prototyping, selling and implementing disruptive storage and virtualization technologies. Prior to Scale Computing, Conboy held positions at Lefthand Networks, ADIC, CreekPath Systems, Sun Microsystems and Spectra Logic. Conboy is notably one of the first movers in the X86/X64 hyperconvergence space, and one of the first 30 people ever certified by SNIA.