31 May 2024, by Slade Baylis
With the ever-growing trend over the last decade being the move of clients from hosting their systems on-premises to the cloud, it was a surprise to most when this trend started to change. With the astronomical rise and adoption of public cloud providers, such as AWS (Amazon Web Services) and Microsoft’s Azure, it would be easy to assume that their offering must have been the best possible option.
What’s slowly dawning on a lot of organisations is that leasing cloud resources has quite a cost, especially when compared to using your own hardware. Something was lost in the transition, specifically how cost-effective using your own hardware can be. Whilst the flexibility of being able to effectively scale “infinitely” was freeing in some respects, most organisations don’t require infinite resources on short notice and have fairly predictable growth that can be planned for.
With most organisations now looking at alternatives, the topic of colocation is more relevant than it has ever been. That’s why this month we’ll be introducing our readers to what colocation is and discussing how our Tier IV accredited data centre makes our colocation services more reliable than anything else out there.
In the late 1990s it was the norm for IT systems to be hosted at the same location as the businesses that used them. The term for this is hosting systems “on-premises” or “on-prem”. For most organisations, the extent of this was simply storing a few servers and networking appliances in an air-conditioned section of their office. For larger businesses and academia, requirements often grew so large that storing these systems - and the infrastructure required to make them reliable - simply became unfeasible.
In response, these larger institutions were able to resolve both their space constraints and infrastructure requirements by investing in their own off-site data centres. For smaller businesses though, the cost to do this was unimaginably high. This problem created a demand for some kind of solution, which ended up coming in the form of the colocation sector.
The idea was relatively simple, build the same type of facility that these larger organisations were creating, but then lease out space in these facilities as required to smaller businesses. This allowed everyone to access the reliability of redundant cooling and power delivery, without needing to invest large amounts of capital required to build, maintain, and physically secure their own data centre.
These days there are several different common approaches to colocation. As always, one of the most common is organisations looking to bring their own hardware from their premises and host it in a data centre. However, the shift away from leased cloud resources has resulted in more people looking to purchase new hardware and even leasing hardware as required.
In many instances, people looking to colocate hardware within a data centre will already own their own servers. This isn’t a surprise, as just like before, many businesses still start out hosting their systems on-site for cost reasons. As their needs grow, such as extra requirements for reliability and security, most will reach a point where they want to relocate their existing hardware and “rack it” within a data centre.
“Racking” hardware refers to mounting hardware within a data centre. This term simply came about from the name of the cabinets that the hardware is stored within, which are called “server racks”.
The process of relocating existing hardware from on-prem to a data centre is commonly referred to as a “lift-and-shift”. This name comes from the fact that, when done correctly, configuration changes can be minimised and it's largely just a matter of physically relocating systems.
As mentioned above, organisations are increasingly looking to minimise their cloud costs. Those looking at non-public cloud options increasingly don't already own existing hardware and so instead look to purchase new systems and colocate them immediately. This approach has gained steam largely alongside the emergence and growth of open-source solutions for running virtualised environments on one’s own hardware. With these open-source platforms, organisations are now able to gain many - if not all - of the advantages they’ve grown to love from public cloud providers.
Those previously exclusive features like: high availability (HA), scalability, virtualisation, containerisation, and software-defined networking, are now all possible on self-managed hardware. The added benefits of a colocated approach, such as having control over hardware from the ground up, can now be achieved without having to sacrifice the reliability or functionality that users have grown accustomed to.
Whilst a hardware-based approach can’t hyper-scale infinitely like a public cloud solution can, most organisations don’t find themselves in the situation of needing to massively scale their infrastructure at the drop of a hat. With this being the case, plus how much cheaper this approach is than leasing the same resources, organisations can now opt to purchase much more hardware capacity than they currently need, and still end up saving compared to public cloud options.
There still exists one downside to owning hardware, which is that of senescence. As hardware ages, not only does it begin to underperform, but it also runs an ever-increasing risk of failing and causing your systems to go offline. This is why it’s recommended to replace systems as they age, either utilising the now-older hardware for non-production/non-mission-critical purposes or selling it entirely. However, there is an alternative approach to this, which is to look at not owning the hardware at all.
An increasingly preferred approach to colocating owned hardware at data centres is to look at leasing hardware instead. As sort of a middle ground between a public cloud and ownership, leasing servers allows you to reduce cloud costs in much the same way as hardware ownership, but without the disadvantage of being tied too closely to the hardware itself. When older servers and systems need to be replaced, either for performance or reliability reasons, organisations can choose to lease new hardware and migrate their systems across to them.
It's due to the flexibility of this approach that many are choosing this path, as it offers a nice balance between reliability, functionality, control, and cost.
If you have any questions about our colocation services, or purchasing and leasing hardware, let us know! We’re happy to answer any questions you have and put together a solution that works for you, whatever your current scale.
You can contact us via phone on 1300 769 972 (Option #1) or alternatively reach out to us via email at sales@micron21.com