Cloud Brokerage Makes It Easier Than Ever To Implement The Rule Of Two
In IT, as in life generally, the rule of two is a useful precept to live by. In a nutshell, the rule of two states that if you only have one of something, you might as well not have any of it. It’s a simple way of expressing the idea that redundancy is the key to reliability. If you’re limited to one instance of something that’s vital to your happiness, efficiency, or productivity, then you’re always only one unforeseen event away from disaster.
The typical example of the application of the rule of two is backups. A common refrain in the IT industry is that if data exists in only one place, its existence is tenuous at best. Backups exist to ensure that there is no single point of failure in a business’ data storage strategy. No backups is akin to being an incident away from annihilation. One backup is great. Two backups — each insulated from events that might impact the other — are ideal.
But the rule of two applies to more than just backups — it can be applied to all levels of infrastructure management — from data storage to servers to entire networks. If any part of an application’s functioning depends on a single server, then a failure in that server will take the application down, regardless of the robustness of redundancy planning in other areas of the application’s infrastructure. You can have redundant web servers, but if those web servers depend on a single database server, then availability is entirely dependent on that single point of failure.
One of the great strengths of cloud platforms is that they make it much easier — and less expensive — to implement the rule of two. Redundant servers can be waiting to take over from nodes that fail. Load balancing across multiple servers is almost as easy. In the cloud, the rule of two can be implemented programmatically — new servers can be spun up automatically in response to server failures. And all at a lower cost than a bare metal implementation of robust redundancy, where the servers have to be bought and maintained in a data center — sitting idle until their moment to shine.
But, as I said earlier, it’s ideal to implement the rule of two across all infrastructure levels, and even in the cloud that hits a snag. On most cloud platforms it’s possible to create redundancy at the server level, the network level, and even the data center level by deploying infrastructure into different regions. But at that point we hit the ceiling. Platforms cannot provide their own redundant instance. And that’s a problem, because it’s risky for any business to rely on a single vendor’s platform.
Platform redundancy has typically not been easy even in the cloud, but in recent years cloud brokerage platforms have risen to make implementing even this level of redundancy straightforward. A cloud integration layer can abstract from the management interfaces of multiple cloud vendors, often spread across the globe, and provide a single point of control and billing — deploying redundant servers on multiple cloud platforms becomes almost as easy as deploying them on a single network.
Cloud brokerage empowers business to apply the rule of two at every level of their business’s infrastructure deployment, ensuring a more redundant architecture that leverages normalized services which don’t require the added complexity of additional API integration.