Matthew Hardman, Chief Technology Officer, APAC, Hitachi Vantara, says the idea of a hybrid multi-cloud is ‘not so wild’ anymore.
In the rapid pursuit of competitive advantage and a favorable position as an innovator, many customers have been led to believe that innovation happens by being in the cloud.
That is not the case.
Innovation happens when you identify a different or improved approach to engaging a customer – or a way to improve the efficiency of how something is done internally. Innovation happens around us every day in varying ways. So, businesses need to have a broader view on technologies that help achieve improved outcomes.
I don’t want to downplay what the cloud can actually offer; in fact, I think the rate that it enables an organization to quickly provision and utilize technology is awesome. It allows organisations to leverage a new database technology and simplify analytics deployment at speed. Cloud is a platform of services and tools you can use to achieve an outcome, but it shouldn’t be considered the end game or a business strategy.
As we have seen the utilization of cloud services accelerate, so too have we seen the acceleration of regulations about how a company’s data is being handled and used in those services, the readiness of skills to operate and secure and how best to manage costs. The idea of a hybrid multi-cloud has been floated recently, driven by a desire for customers to take back control and protect data, manage costs and offer flexibility in choice of services. The promise of a hybrid multi-cloud is an architectural approach where you separate the data from the services being utilized in the cloud. This approach separates the data from the cloud services, enabling you to control the data in your owned infrastructure, while enabling a choice of services across clouds without incurring the costs of moving data out of the cloud – commonly referred to as egress costs.
Many organisations are currently reassessing where they can host their applications and data. While the immediate idea might be to bring it back to data centers and private clouds on premise, there are other options available to customers. Customers might consider the use of colocation facilities, that is utilizing a shared data center providing customers with floor space, cooling, connectivity and more – in this case it is an extension of the customers’ data center as they own the hardware.
The unique advantage for some of these colocation providers is that they also host many of the cloud services being used, so direct connectivity can be provisioned. This ensures that the performance of the application will not be impacted, should the data be hosted externally to the cloud services or in a colocation.
A key Driver of this shift is due to what customers are experiencing regarding cost. The original migration towards the cloud was certainly something that was touted by significant marketing campaigns with the promise of limitless resources, available on demand, scalability, flexibility and productivity.
However due to limitless scale and access to services, some applications basically overdid it, they over architected, over provisioned and replicated data in different areas.
This led to an explosion of uncontrolled costs.
A great example of this is David Heinemeier from the company 37 signals, where they went from spending $3.2 million a year on cloud to investing in an on-premises infrastructure for $600,000 hosted at a colocation provider – thereby saving a million dollars on infrastructure.
An IDC study found that more than 70% of enterprise customers are starting to repatriate their workloads. The reason being is that they have been affected by issues of costs, regulations and security issues.
Local government authorities play an important role in guiding institutions on the use of a variety of technologies in their day-to-day operations – and cloud technologies are certainly a key area of focus.
While many of these regulations do not discourage the use of cloud technologies, they do outline an approach to ensure resiliency in the operations of the institutions they govern.
As an example, countries like Singapore and Hong Kong encourage financial institutions to have a strategy in place that ensures mobility of their applications and data in times of operational issues and or breach of security.
Recently, the Hitachi Vantara Modern Data Infrastructure Dynamics Report revealed six out of 10 companies in Asia are overwhelmed by the amount of data they manage and 73% are concerned their current infrastructure will be unable to scale to meet upcoming demands, leaving them exposed to security challenges.
Businesses have a two-year window before data needs are nearly expected to double. Applying these same factors of growth to cloud based data only further exacerbates the challenges on data mobility.
Embracing a hybrid multi-cloud approach enables customers to better manage a strategy of mobility, cloud to cloud or with on-premise resources.
The idea of a “hybrid multi-cloud” is not so wild, in fact just like core application development techniques, we would architecturally break up an application into data, services and client, and so in this approach, we are just geographically breaking up the application. Your data is placed in a centralized location not in the cloud, but near-cloud, with direct access to data processing services in the cloud. You get the benefits of control, protection, and security of your data, with the flexibility in choice of cloud.
While hybrid multi-cloud has become the new formula, I’d like to take it one step further and suggest that the solution lies in hybrid platforms, an architecture that truly encompasses cloud, co-located services and on-premise. This could also include the distributed cloud that locates cloud services within the boundaries of their own data centers such as Google Anthos or Azure Stack HCI.
Cloud repatriation is happening as companies are taking back control of their data. The technology is here – so there’s no holding back.
Click below to share this article