Predictable VPS: what businesses are actually paying for, and how fair virtualisation works
A VPS is almost always described as a more powerful package – and there is some truth in this; the specifications do indeed show higher figures: more CPU, more RAM. But what really determines whether a website can handle the load on a VPS has almost nothing to do with these figures.
The amount of resources in itself is only half the story. The crux of the matter lies in how they are allocated. In theory, virtualisation works perfectly: the hypervisor allocates a quota of processor time to each virtual machine – a guaranteed minimum number of clock cycles on the physical cores. But back in the real world, businesses regularly find that a website on a VPS still slows down due to other users’ traffic.
In the following posts, we’ll discuss in detail the hidden limitations of the mass market that undermine the hypervisor’s mechanism, and why cheap computing power ends up costing businesses dearly. But one thing is already clear: the solution to this problem always boils down to one thing – moving away from illusory figures in favour of genuine hardware isolation.
The market and marketing have conditioned us to buy gigabytes on screen, forgetting that behind them lies the physical economics of data centres. When a provider promises huge volumes for a couple of dollars, this is by no means a gesture of goodwill from a ‘good Samaritan’, nor is it a strategy to maximise profits from mass sales. More often than not, it is a way of compensating for a lack of redundancy and the use of cheap components. At best.
We at Onhosted know this side of the market all too well, having ourselves gone through the process of trying to stabilise projects in an unpredictable, third-party environment. Tired of struggling with infrastructure that operates according to its own hidden rules, we decided to build a platform from the ground up – focusing on true isolation. We have designed two VPS/VDS ranges so that fundamental reliability (fair resource allocation and RAID 10 disk mirroring) is built into both, but you only pay for the architecture that your product actually needs:
Simple: A range designed for personal projects, test environments and non-mission-critical business infrastructure. We ensure a balanced allocation of resources across high-quality hardware. This is a predictable environment for stable operation, where reliability is not sacrificed for the sake of availability.
Business: A solution for resource-intensive systems and high-load services, where consistent performance is essential and downtime is unacceptable. The platform runs on high-frequency processors and enterprise-grade storage. Here, we incorporate data redundancy for key components to ensure stable operation even under extreme loads.
When it comes to complete physical independence, this is the realm of Dedicated Servers, which projects typically move to at the stage of global scaling. However, for most day-to-day tasks, a properly designed VPS provides the necessary foundation: resources can be scaled up without downtime, root access allows you to build a custom network architecture, and fair resource allocation ensures that someone else’s problem won’t become yours.
Any claims regarding fault tolerance must be put to the test in practice. That is why our architecture is backed by a strict SLA, and we offer you a 7-day full trial. Visit the Onhosted website to choose a configuration and start your trial
