What hybrid cloud solutions generally have in common is the fact that multiple resources are managed from a single software platform. An example of this is how Microsoft Exchange and Exchange online can co-exist and are managed from a single utility. However, it’s worth noting that, as is often the case with cloud definitions, everyone has their own interpretation, so you’ll definitely see and hear variations on this theme.
Hybrid cloud involves pairing up a pair of different environments (let’s say a public and a private cloud) using a layer of software, such as VMware Cloud Assembly, to share data and manage/ orchestrate the two environments. Multi-cloud involves two different IaaS providers (like Azure and AWS) but without the same requirement for data sharing or orchestration.
An easy way to think about the two is that, with hybrid, you could manage one workload over two or more cloud platforms, whereas multi-cloud is more likely to be separate workloads on separate platforms. Multi-cloud is somewhat broader than hybrid – you could have a hybrid cloud within a multi-cloud… but not vice versa.
Why would you want to host workloads on entirely different cloud platforms or with different providers? Well, there are quite a few reasons…
Hybrid cloud architectures can be used as an organisation’s first step into the cloud, extending their on-premise or own data centre resource with the addition of some public cloud consumption. This allows businesses to extend their capability, make the most of their current investments, improve resilience and move applications closer to end users.
It’s been the talk of analyst reports for several years now, with many expecting a widespread shift to multi-cloud over the next few years, but what makes it so appealing? Well, as explained by Cloudhelix Cloud Architect, Ed Breslin, you might not even realise you’re a multi-cloud organisation:
“Web-based email as a service is so common now (and arguably pre-dates the 'cloud' jargon) that an organisation using Gmail for email and AWS for some servers is already multi-cloud, but they may not have noticed!” Ed Breslin – Cloudhelix
Each public cloud platform has its strengths and specialities, which means that for businesses who are big on public cloud, there’s a real benefit in getting the best out of each public cloud platform across their ecosystem.
Investing heavily in one provider is seen by some organisations as a risk, both in terms of vendor lock-in and availability, which results in a multi-cloud, as operating more platforms can allow this risk to be reduced. As much a business continuity measure as an IT one, some organisations may choose to work mainly in, let’s say Azure, but leave a skeletal pilot light setup in AWS, in case Microsoft have a huge outage. Outside of this, some businesses feel comfortable working mainly in one platform or technology while some may prefer to spread this out.
Private and dedicated platforms offer real flexibility and customisation, making them more suitable for intensive and business-critical workloads. These platforms give organisations greater control at various levels of the infrastructure stack when compared to a public cloud, which is why less business critical applications may be are migrated away from private cloud infrastructure to public, which also helps to ensure there’s enough resource available for critical applications on private and dedicated environments that could be more costly to run.
There are tools out there that assist with the management of multi-cloud, adding a layer of abstraction between cloud platforms to allow management, monitoring and analysis. However, throwing tools at a problem won’t solve them alone – if the move to multi-cloud wasn’t effectively thought through and architected, a layer over the top isn’t going to make headaches and frustrations go away. A well architected multi-cloud solution may well be easier to manage than a poorly setup hybrid solution, which can also be assisted by an organisation’s willingness to embrace managed services rather than servers.
As this is a growing and relatively new space, the tools themselves are new and engineers who know the ins and outs of them might be hard to come by. A solution that provides visibility and metrics for various services across multiple platforms is vital to making informed management decisions. For a project of this size and nature, it could well be worth consulting a provider or external consultant you can trust for part or all of the migration project.
Getting to a place where there’s a range of hosting locations in the mix–and also several cloud platforms–brings added complexity for infrastructure teams, but with this complexity also comes the ability to add greater resilience and drive real value through to the wider business.