Are we ready for the Multi-Cloud revolution?

The Multi-Cloud concept is part of the maelstrom of terms and technological trends that have arisen around the Cloud Computing in recent years: Cloud Brokerage, Cloud Orchestration, Cloud Bursting, etc.

First of all, let's define what is Multi-Cloud. Basically, it would be based on combining Clouds from different providers to serve a single IT platform from a heterogeneous architecture. On paper, our service (website, business application, online app, etc.) and its users will have the best of each CSP in terms of performance, availability, positioning, cost or any other criterion we want to consider.

The Multi-Cloud option should make it easier for the technical team to use a single cloud infrastructure panel to move process loads between one provider to another, according to time slots, prices, overload levels, etc., or that the Multi-Cloud solution itself offers automatic load balancing tools. The theoretical advantages are obvious: more efficiency and better performance for users, who will benefit from the Cloud that provides better quality of service. At the same time, and from the most practical side, the difficulties to implement Multi-Cloud in the vast majority of organisations and companies multiply significantly.

For example, with Multi-Cloud we will have to know the different levels of service from one provider to another, manage and negotiate SLA agreements with each one or take into account the regulatory framework of each provider, including the possible differences that exist in each country regarding personal data processing. In addition, being realistic, it is more than likely that we will find some incompatibilities in the Multi-Cloud panels in order to synchronise the technical features of different Cloud platforms.

For this reason, a Multi-Cloud strategy may not be an affordable option for all companies. Rather, it is an alternative recommended for those organisations with mature Cloud environments, focused on DevOps practice and highly experienced in the use of solutions in the Cloud.

How do we know if we are ready for Multi-Cloud?

Before deploying complex IT solutions such as a Multi-Cloud environment, it is essential to analyse the needs of each company and determine their suitability: competitors, operations in different countries and markets, low latency requirements, application of different regulatory frameworks, etc. All these elements can constitute scenarios in which it can be justified the adoption of these Multi-Cloud solutions, but they are also elements that we must define and clearly identify when choosing a single provider.

On the contrary, if what we are looking for is the step from services on premise to the Cloud model, or successfully undertake a first stage, or even the consolidation, of our Digital Transformation strategy, we could go too fast if we directly opt for Multi- Cloud. We run the risk of falling into a technological framework that, instead of facilitating our tasks, hinders our inevitable migration to the Cloud, fact that will undoubtedly lead us to feel disappointed after all the benefits heard around the Cloud.

Migration to Cloud in five stages

In the adoption of Cloud, it is convenient to be clear about the path and the partners we select, but also the intermediate stages. Undertaking the journey by establishing well-defined goals can mean the difference between success and failure in Cloud adoption.

For example, before implementing Multi-Cloud or any other future trend that may arise around the Cloud, we can define the following milestones or intermediate stages:

  • First step. Start with a non-critical area with which the IT staff can experiment the different possibilities of a Cloud platform. You can even manage a mixed environment, where basic services for the business, such as management applications, are kept internally, and others such as the website or the backup go to a Cloud environment.
  • Second step. Combining services deployed in Private Cloud with other deployed in Public Cloud, from a single panel that allows us to move loads between both environments when we need it and within the scope of relationship provided by a single provider, end-to-end responsible for the entire service.
  • Third step. Use hybrid environments, where Cloud infrastructure solutions coexist with physical environments, combining the benefits of a common infrastructure on demand, and the possibility of having back-ends and their own applications in a robust and dedicated infrastructure.
  • Fourth step. From the same control panel, under the same contract and with a single service provider, have the possibility to deploy infrastructure in different Cloud nodes located all over the world.
  • Fifth step. Deploy the necessary monitoring and implementation options on Cloud infrastructures, including the required automation to orchestrate load balancing and redeployment to minimise response times and economic and human resources allocated to such tasks.

As it usually happens in the technological universe, there are no silver bullets and it is also sensible to bear in mind that these different stages do not have to get rid of the services deployed on premise, with which it is possible to coexist perfectly.

In the same way, it is not compulsory to follow a sequential order from one stage to another; any of these steps can abandon its temporary nature and consolidate itself as the optimal scenario for an organisation in its adoption to the Cloud.