Exploding cloud myths, part 3: Why the job is not finished once you have migrated
Moving to the cloud is not called a journey for nothing. Cloud journeys, like other transformations, comprise a never-ending bell curve of change.
So, while it may be tempting to crack open the champagne once you have migrated some initial services to the cloud and shift attention on the next item on the CIO’s ‘to do’ list, the job is far from over. This period simply marks the point at which the cloud journey veers off on a different tangent.
A business-driven transformation
Because moving to a cloud model is a business-driven transformation, not a technology-driven change in service supplier: the journey is essentially a chain of initiatives stretching out ahead of you, only the earliest of which can be clearly defined today.
Outsourcing is often a key driver for a move to the cloud. In these cases, the primary goal is to move services to external providers with specialist expertise who can deliver the services for a lower cost, while agility and flexibility remain collateral benefits. These migration activities can be readily encapsulated within a specific project, with the timescales, budgets and scope being agreed by all parties.
In theory, going live using the new technologies/suppliers will mark the project conclusion, and present an opportunity to measure the value delivered, followed by the chance to celebrate accordingly.
Moving the cloud goalposts
The reality tends to be rather different.
During the early stages of a project, the team becomes increasingly aware of the agility, flexibility and modularity made possible by the cloud model, leading to an expansion of the original strategy. What seemed like a finite goal at the outset morphs into something bigger.
This can be attributed to awareness of the ‘art of the possible’ and a realisation that these characteristics can help the business keep pace with, or overtake, competitors. This explains why what starts as a migration project can, if care is not taken, grow arms and legs - and may ultimately become as unwieldy as an octopus.
In just the last few years, virtualisation has evolved the physical allocation of hardware resources between operating system instances to be software-controlled and more modular as a result. Containerisation built on this, evolving from operating system instances to application environment instances, sharing binaries and software libraries. Cloud has utilised each stage of this evolution. We are now seeing: ‘containers as a service’, which moves traditional ‘infrastructure as a service’ closer to ‘platform as a service’ for consumers; and serverless computing, which is exploiting microservices architecture (itself an evolution of service-oriented architecture) to offer ‘software functions as a service’.
The beauty of the cloud model is that it can adapt. But organisations still need to retain control over any strategy change. This can be achieved most simply by sticking to the original drivers and benefit measures, while starting another initiative in parallel to address the new drivers.
While organisations will undoubtedly generate value with this syndicated approach to cloud adoption, more value can be more readily achieved by thinking further ahead and establishing at the outset a longer-term transformation plan.
Organisations taking this longer-term, more strategic approach will already have plans in place to deliver agility and flexibility within their operating model, and to exploit modularity in architectural and commercial management. That said, of course nobody can truly predict how cloud architectures will evolve beyond the next few years.
Opinions tend to be driven from two different perspectives: the software teams with their focus on cloud-native application development; and the infrastructure teams with their focus on platforms, virtualisation and containerisation. Whatever prevails, there is little doubt that these two perspectives are heading towards common ground.
Legacy looms large
But both face a significant and unavoidable enemy: supporting, and integrating with, legacy technologies. After all, new solutions have emerged over decades, and will continue to do so. Today’s innovation is tomorrow’s legacy.
The only sustainable strategy to tackle legacy is to do it head-on: in fact many organisations are finding they need to do this before cloud migration even becomes feasible. Next they need to put in place comprehensive and persistent technology lifecycle management practices. This is something that organisations will have less control over with cloud provision in any case, so becomes a prerequisite.
Even without legacy considerations, one of the key benefits of a cloud model is the flexibility in service provision. Many organisations that start out with an outsourcing agenda neglect the supplier switching option and inadvertently end up being locked into a vendor. This may be through dependence on specific technical features, by limiting staff skills, or by simply not considering any future desire to utilise another supplier (e.g. for improved service features, resilience or costs).
It goes without saying that building-in a multi-cloud strategy from the outset can avoid considerable pain in reverse engineering further down the line.
Cloud as a moving target
The extent to which organisations wish to invest in their own IT (and how they define what this means) versus using IT services from specialist providers, will continue to vary in different ways as the industry evolves. The same applies to cloud.
Cloud started as a means to support IT agility and modularity, but has become a reincarnation of outsourcing - and is rapidly becoming the next-generation platform for software development. As organisations exploit different benefits, they will need to adapt their business case measures and key performance indicators accordingly.
In an uncertain world, the only certainty is that evolution is here to stay, not just within cloud itself, but also in the ways in which we exploit the different potential benefits this presents. Enterprise cloud adoption is still at an early stage in what will be a long journey. Belt up for the long haul.
Interested in hearing industry leaders discuss subjects like this and sharing their experiences and use-cases? Attend the Cyber Security & Cloud Expo World Series with upcoming events in Silicon Valley, London and Amsterdam to learn more.
- » Sainsbury’s looks to Google Cloud for machine learning as retail cloud case studies continue to climb
- » Amazon completes consumer database migration from Oracle to AWS
- » Building confidence and power: Exploring greater female leadership and participation in cloud and data analytics
- » Best security practices for migrating to the cloud: A guide
- » Is performance engineering still needed when it comes to cloud?