To ensure the IT requirements of your business are appropriately met, your choice of server – whether a dedicated physical, cloud-based or hybrid solution – represents one of the most significant decisions you can make from a technological and operational perspective.
It is a particularly important decision when you consider your server is the means by which your mission-critical business applications are housed and accessed.
The higher the levels of uptime required from a specific application, the higher fault-tolerance the hosting environment needs to be able to demonstrate. Before virtualisation and hardware consolidation, your mission-critical applications would have typically been facilitated by dedicated, on-site equipment. In today’s increasingly complex operating environments, a costly investment in a purely physical server may no longer by enough to facilitate the high levels of usability and uptime businesses expect.
Every company varies in its IT needs. There are many combinations of hardware and virtualised elements that can be employed to create your ideal server solution, meaning lots of considerations need to be made to ensure you enjoy peak performance from your critical applications.
An increasingly popular choice is a cloud-based dedicated server. This is essentially a hybrid solution that offers the processing power enjoyed by users of physical servers, and the cost-savings and scalability appreciated by advocates of the cloud. The limitations of either a single, on-site dedicated server or an entirely cloud-based solution can be overcome by using the right combination of expertly managed hardware, and a virtualised environment.
By incorporating at least some elements of virtualisation into the way a business stores and accesses its mission-critical applications, adopters stand to benefit from a reduced hardware footprint, lower data centre costs, and improved ROI.
So how does it work? Put simply, a company can have multiple applications running on a single physical server that plays host to several virtual machines (VMs), eliminating the need for individual pieces of hardware to host each application. And while dedicated servers have maintained an edge over virtual alternatives when it comes to latency, as the sole user of this hybrid set-up you won’t have to compete for bandwidth when accessing your virtualised applications, meaning none of the frustrating delays often associated with the public cloud.
Plan, plan, plan
Planning ahead when selecting the right server option for your business is vital for identifying and maximising hardware and virtualisation opportunities. Your ultimate solution will consider the unique needs of your business, and will have the potential to enable improved reliability, flexibility, scalability and efficiencies.
Before making a decision on your ideal server, spend some time identifying which of your existing applications can be considered mission-critical, or are likely to become so in the future.
Once you have successfully outlined what a mission-critical application looks like for you, you can begin to determine the most appropriate means of delivery, based on the level of server resources these applications are likely to require. Some factors to consider include:
- the goal of the application
- the number of users
- the resources currently allocated to the application
- the resources the application is expected to require in six months, one year, and beyond …
Understanding the application’s current goal and future requirements will help determine what you need from your server in the immediate, short and long-term. You may choose to research the options yourself using advice available online, or you might consider engaging a system integrator or consultant to look at your current architecture and make smart recommendations for improvements.
Performance is key
Ensuring the necessary resources are available to power your mission-critical applications is crucial if you are hoping to enjoy measurable performance gains and improved productivity.
As outlined above, considering and understanding the potential increased demands on a critical application in the future is important. If growth is on the cards for your business, you’ll need to make sure your chosen server can offer you the scalability required to accommodate this. Anticipating an application’s future requirements will equip you with valuable insights to inform smart decision making around your server.
Virtualisation: The barriers and advantages
There is no denying that virtualisation has become an increasingly common solution to support business operations. However, there still exists some reluctance amongst both developers and business owners to utilise a virtual environment for applications that require high levels of uptime. This makes sense when you consider these applications are often the most complex, and pose the biggest risk if disrupted. Naturally this leads to concerns around application effectiveness in a cloud environment, security and compliance, and stakeholder buy-in.
The good news for cloud enthusiasts is the increasing number of vendors who are confident and comfortable with the idea of deploying their applications using a virtual platform. Big-name software developers like Microsoft have demonstrated a commitment to creating virtualisation-friendly programs and applications, with many now optimised specifically to operate in this kind of environment.
Similarly, physical server hardware now often comes virtualisation-ready, with manufactures proudly touting impressive metrics to demonstrate how well their products perform in both hosted and physical on-site environments.
While the cloud itself is generally very secure, deployment of workloads and the way the cloud is utilised by the individual business can result in certain vulnerabilities. Your unique network set-up will require you to consider the security measures that are essential for you, ranging from secure cloud-to-cloud connections, to virtual firewalls. An expert provider or consultant will have the ability to anticipate security and compliance hurdles, and provide recommendations to ensure all necessary measures are put in place to protect your server from breeches.
Additional benefits of virtualisation include improved redundancy thanks to ‘VM failover’. Essentially, this is the ability to mirror the VM at an alternative hosting location many kilometres away. In the event of disaster therefore, your critical applications will continue to run seamlessly.
Hardware and storage
Any decision relating to the most appropriate use of server hardware to support your critical applications must consider factors ranging from flexibility to migration, consolidation, networking functionality and storage.
When it comes to on-site storage, some may opt for a more traditional ‘rack mount’ system, which houses larger network servers and their associated devices in a protected and controlled environment. Rack mounting has its advantages, including easy access to the internal elements of larger physical servers, without the need to remove the entire server from the rack.
Others prefer a ‘blade environment’, in which the server is stripped-back with a modular design to minimise the use of physical space and reduce the amount of energy required to keep it running.
A focus on redundancy
A fault-tolerant system is vital to ensure zero downtime for mission-critical applications and optimal business continuity. If the physical host of these applications were to fail with no precautions in place, the impact on your business could be catastrophic. Therefore, it is essential that any server operating as a physical VM host is fortified with redundant components and primed for VM failover. Data duplication, snapshots taken at set intervals, and offsite physical server mirroring each play an important role in supporting the swift recovery from system failure, and keeping your applications functional and accessible. In particular, elements including power supply, hard drives, random access memory (RAM) and network interface cards (NICs) should all be made redundant as a priority.
Recovery time objectives
The impact of a disrupted critical application varies greatly depending on a multitude of factors, including its function, and the specifics of your business operations.
Applications identified as mission-critical should utilise a server that offers a set recovery time objective (RTO). This is essentially the maximum time between disaster and recovery, and is designed to minimise disruption to your operations and allow for business continuity to be maintained.
An appropriate risk assessment process can help inform the parameters of your RTO.
Remember to consider memory!
Whether you host your critical applications in a virtualised environment, use a dedicated physical server, or opt for a hybrid solution, memory has an important impact on performance. The term ‘mission-critical’ implies ‘memory-intensive’. While virtualisation offers limitless opportunities for ramping up memory, todays hardware also tends to come equipped with significantly more memory capacity when compared to previously available equivalents.
When selecting a server, the smart thing to do is to verify the expansion capabilities of the technologies involved, while also researching the level of RAM it can support.
Designing a server environment with open standards in mind will stand you in good stead for future success. Open standards are developed and maintained through a collaborative and consensus-driven process, and are publically available. They serve an important purpose by helping to enable compatibility amongst lots of different products.
Having a server built to open standards supports ultimate flexibility, which is important when you consider how rapidly businesses and their IT requirements are evolving. These standards allow you to access a wider range of vendors when it comes to implementing any future server upgrades required to keep your critical applications functioning properly.
Try before you buy?
Often we don’t know exactly what we need until we need it. Server design, purchase and deployment can be a costly investment, and one you should only make once you are entirely confident you have found the right solution for your business. Talk to your vendors about a Proof of Concept (POC) or advanced demo. This will allow you to see in practice how effectively the chosen server set-up will fulfil the requirements of your critical applications, before you make a final commitment.
This process can also provide a valuable opportunity for engineers to eliminate any problematic variables before the technology goes live, ensuring your new server is primed to offer the best outcomes for you and your business.