As more and more businesses address their operational and technical needs in the wake of COVID-19 – and all the challenges it has brought – an increasing number of enterprise workloads have been moving into different clouds.
However, according to a study commissioned by IBM with McKinsey & Company at the end of 2018, 94% of enterprises have a multicloud strategy and are either actively doing cloud-native development and deployments, or just have traditional cloud-based workloads.
Even though the trend to cloud is clear, many businesses running cloud technology have only migrated 20% of their actual workloads to a public cloud– meaning the vast majority of existing workloads, one could say ‘legacy’ in light of the cloud-native trend, currently reside in on-premise [virtual] servers. In other words, hybrid cloud management will be the predominant management model for years to come.
What are some of the challenges they face? How can companies continue their migration – and why should they?
The Agile Approach
You often hear words such as ‘agility’ and ‘flexibility’ in the cloud and cloud-native industry. And while some view them as buzzwords, they have a solid foundation of meaning behind them.
Take ‘agility’ for example: Imagine a developer has a resource profile for an application (CPU, memory, disk, and network). That developer needs those resources fast. But for them to be available on-premise, their company must invest serious money for CAPEX.
However, when you consider that these resources will only be used, perhaps 20%-40% of the time, the investment becomes a questionable use of capital. Because, when people leave the office in the afternoon, those bought resources will still be active – but are likely not being used and cost money – just ask the CFO about asset amortization!
While this is arguably stifling innovation, from an economic standpoint, why should companies pay millions of dollars for servers that do nothing for 8-16 hours a day?
This is precisely why Amazon started its Web Services (AWS) in 2006. With massive data centers – from their online business and their servers – doing nothing most of the time, the decision was made to sell excess capacity: resulting in EC2, the first public cloud.
A new factor not even conceived with the advent of cloud-native computing or even considered as recently as the referenced IBM study with McKinsey, is the Work From Home (WFH) revolution, triggered by the breakout and proliferation of the COVID-19 pandemic. There is now in many cases, no office – and in parallel the public cloud data centers have become more prominent in new IT strategies. This ‘cloud-first’ policy for development will accelerate cloud-native development adoption.
All of these issues underpin the benefits of [public] cloud computing. Businesses have now more need to make use of these services, and those with less need also require access to them to gain the benefits cloud computing affords their business commercially and operationally. In other words, they need to be able to pay for what they use, scale up when needed, and deploy individual services directly – all within technology and financial boundaries set by corporate management with auditing and governance capabilities.
This is why cloud management platforms (like CloudController) are needed – for both public or on-premise cloud service orchestration, deployment and usage controls.
Reasons Why Companies Wait
So if adopting a cloud-first policy is essential, and this is one of the requirements for a digital transformation strategy, why have they only migrated 20% of their application workloads? After all, 94% of companies have taken to the idea of cloud-first – and are already developing to and using public cloud services!
There are two main aspects to it. Firstly, on an operational level, transformation means that jobs and business processes will change. A software developer’s day, for example, will look very different both during and after a migration journey – certainly if the application(s) they’re responsible for developing and maintaining need to be rewritten to be migrated to a public cloud. This change can be reason enough for a company not to embrace it immediately.
Secondly, on the technological side, even with all the fantastic technology at their disposal – like containerization, microservices architectures, DevOps automation, multi-tier platform rollout, and cloud-native – some companies are afraid of making drastic changes to their traditional ways. You can’t just simply ‘move’ some applications. Applications that run mission critical processes for major corporations cannot just be made cloud-native microservices-based in a flash. They need to be rewritten and/or redeployed with new technologies, etc. And to complicate matters, we’ve seen that even new core technologies like Kubernetes for container orchestration can differ between the ‘standard’ open source version and the version used at individual public cloud providers. So much for the 100% interoperability argument.
Thus, as one can easily surmise, the fear is not always budget-related – they just want their choices to be smart.
Applications need to work!
Companies that are apprehensive about security will also delay migration. They might be wary of digital transformations and seemingly “out-of-sight” models, or their culture tends to wait for others to be early adopters, thus they prefer to wait until more companies do it first. As a strategic salesman I can’t tell you how many times I’ve heard a prospect say, “How many references do you have in [insert their country name here]?”.
For a good analogy to this, remember, a generation ago, enterprise data centers migrated en masse away from mainframes to client/server architectures, yet still today mainframes are around and are running mission critical applications. The mainframe market still grows year-over-year at about 4% CAGR. [Source: Allied Market Research]. In other words, some workloads will continue to run on-premise for the foreseeable future.
However, few people question the security of a cloud anymore. And the many advantages it provides make it well-worth the migration efforts. The truth is, adopting cloud is now an essential part of IT strategies at all enterprises.
The Need For A Cloud Management Platform
Looking ahead, based on current trends, 5-10 years from now, 80%+ of application resources – read: infrastructure – will either be resident at cloud providers, or the applications themselves will be cloud-native. A uniform management plane will take a more prominent role in both deployment and management. Companies that resist will struggle tremendously. So this is why there’s such an urgency businesses to bridge this ‘hybrid cloud gap’ now – and why tools like CloudController are needed.
CloudController doesn’t change your core technology choices – it enhances them. When your environment’s setup is complete, our Cloud Management Platform automates how everything works – ensuring a true “Single-Pane-of-Glass” experience from the interface in which you can execute all your cloud management, monitoring, and control functions, such as:
- Control how virtualized on-premise data center assets are used and by whom
- Deploy and manage from AWS, Azure, IBM Cloud or GCE (if public clouds are used)
- Container orchestration environment integration and automation
- Cost insights and associated controls
- Synchronization with your ITSM platform
No matter what core technology you’ve chosen for building platforms, containers, or virtual infrastructure management, a cloud management tool is needed for managing and orchestrating service deployments, governance and keeping costs in check – across multiple resources, both in clouds, and on-premise.
Overall, while most companies start small on their migration journey, if they commit to getting 80% of their application infrastructure into a cloud lifecycle, they could save significant money every year in operational costs, benefit from scalable resources, and enjoy best-in-class cyber security.
Inevitably, resistance (as they say) is futile.
Is your company ready to embrace complete Hybrid Cloud automation, orchestration and control? Contact the InContinuum team today for more information about CloudController.