The cloud hosts cloud is high performance, dynamic, scalable, resilient and flexible.
That’s easy to say, but where is the proof in the pudding and what is it made of? In a series of blogs we will look to explain some of the elements and technologies, some familiar, some maybe not so much, involved in producing a solution with so much to offer – and so much to live up to!
To enable the discussion and description of how these 5 areas are achieved we will delve through varying technologies and equipment in use within the cloud environment to piece together what makes it tick.
Firstly, in an effort to focus on resilience, we’ll talk about one of the cornerstones of the cloud environment. Shared Storage.
Shared Storage is fundamentally one of the most important technologies in the environment. It is typically formed by several devices somewhat akin to servers which are filled with hard disks. The devices aren’t ‘just a bunch of disks’ however, and have significant intelligence built in along with resilient power supplies and processing units.
These devices can be connected in various ways but the single most important thing about them is that multiple servers can connect to them and access the same data stored upon them. The important thing here is that should a server fail, the storage doesn’t.
Not at all obvious to begin with, the benefit of this is that if you put some other clever technology or ‘logic’ on top of this you get what is known as failover.
To explain this in a little more detail requires the scene setting so, for this example in the case of the cloud hosts cloud, we’ll deploy a standard specification cloud server on to this environment: 40GB data storage, 1GB memory, 2 vCPU’s (virtual processors) and a network connection.
So, that was easy wasn’t it! Joking aside, this single server now online on the cloudhosts cloud isn’t just a single server; it’s much more than that. Whilst it may be limited – as all servers are – by its specifications (memory, storage, processors etc) this server has one thing about it that a standalone physical server or computer (your PC for example) doesn’t have – Failover.
It is a fact of life that hardware will inevitably fail – disks, power supplies and most commonly anything with moving parts. This means your websites, services, email etc fail and stay off until they are fixed.
This is not the case with the cloud. Failover, built on the technology discussed allows for the cloud server to resume service elsewhere. The cloud server its data and its specification are stored on the Shared storage, a physical server accesses the data on the shared storage and hosts the cloud server.
To enable the cloud server to run, the physical server creates a ‘container’ or bubble in which the processing for the cloud server occurs. In the case that the physical server loses power, has a disk failure or another inevitable hardware failure – the container is automatically and immediately recreated on another physical server in the infrastructure and returned to service.
Thus the term failover and the enhanced ability of the cloud environment to suffer failures of equipment – even if you do only buy one server.
Next time – how the technology enables performance on this environment.
p.s. you may be thinking that this ‘shared storage’ is capable of failing and hence negating all of these benefits. And whilst it is true that anything is capable of failing, Shared Storage is actually not just one device, it can be built in many ways but in the example of cloudhosts infrastructure the shared storage is actually many independent devices each with their own built in failover. This allows for the same benefits at the storage level as discussed earlier. We will go in to more detail on this in a future post – please leave comments if you would like that sooner rather than later!