Why move applications and IT to the Cloud?
Firms have a Business Strategy. IT must support that strategy. Within IT there should be a Data Strategy. The Data Strategy must support the Business Strategy. Indeed, it is increasingly clear that Data Strategy is a large part of Business Strategy. In this regard Data must be clean, useful, available and relevant to support the Business Strategy. Any usage of the Cloud must answer important questions:
- What is our Business Strategy and how does the Cloud support it?
- What exactly in the ‘Cloud’ do we need to use and why?
- How much will the Cloud system or systems cost, and what is the ROI?
- How clean, good, useful is our data and where is it and does it have integrity?
In our view, Data Strategy and understanding the quality of your Data is the first and main step in moving to Cloud computing. This includes documentation, modelling and Data cleansing.
- How is the Business Strategy supported by Cloud Computing?
There are many reasons to invoke Cloud computing including:
-Reducing IT costs
-Increasing IT flexibility and reliability
-Outsourcing IT which may not be a core competency
-Transform business processes through the adoption, and customization, of Cloud architectures and systems.
The key issue is to understand in detail, what exactly within Cloud Computing you want to use and why. Some examples may include:
-Reducing our own data centre costs by renting servers and platforms
-Improved capacity and bandwidth for our applications and systems
-Mobility, and the ability for end users to access data from anywhere
-Migrating data applications which cost a lot to maintain and keep alive
-Force our IT systems to be best of breed, and address internal private data centre concerns such as reliability and security
-Generate more revenues through better client experiences.
The specific Cloud systems you will use, will be based on your use case, what problems you are trying to resolve and what ROI you expect from deploying into the Cloud. You must make a coherent and logical Business Case first, when considering a move to the Cloud.
- What is the Cloud and what do we want to use?
Cloud Computing for many, means server virtualization and renting out an infrastructure and perhaps an OS and SDK platform, and maybe some databases. Cloud Computing is however, more varied, complicated and important than mere virtualization. The power of Cloud Computing rests in the transformative nature of the services, technologies and stacks available, and employing best of breed practices. Cloud enforces the very best IT methods, including security.
By understanding what problems you want to resolve you can decide on what exactly the Cloud, and a future migration path using Cloud services, might look like:
-Simple migration of assets, applications or code bases to the Cloud [eg putting your Exchange servers into O365 – something that can be done in 2 weeks]
-Moving data and applications to the Cloud in order to reduce costs and bottlenecks
-Moving enterprise systems to the Cloud in order to allow mobility and data access from anywhere
-Transforming business systems by mapping distributed Cloud processes to our specific business needs
-Using advanced Cloud services ranging from IOT, to Machine Learning, to dramatically reinvent business processes, products and client-centric services
-Using best of breed IT practices from Agile to Security models to significantly upgrade and improve all IT systems.
The Cloud is not simply about saving money. In the end it should be about business transformation.
- How or What is Our Strategy for Getting There?
Your overall Business, IT and Data Strategy is critical. In order to have a life-cycle roadmap and plan for Cloud usage you need to do the following:
- Cloud options must match the expected Business outcomes and overall strategy
- Consider a hybrid Cloud model which may provide many options for different workloads.
- Cultural and organizational change will occur in both the IT and Business units. Training will be mandatory and needs to be budgeted.
- Do pilot projects and deploy, or migrate small workloads or application pieces to the Cloud. Doing less and understanding the various complexities of the Cloud, will allow you to do more over time.
To ensure ROI, make the metrics concrete and real. Don’t say ‘productivity will improve’. That does not mean anything. Instead the benefits need to be real; staff reduction, call centre reduction, reduced maintenance and IT costs, improved cash flow collection, more orders, more revenues, specific product improvements which lead to more sales.
Cloud Computing should be separated as a core skill
If you have a large IT infrastructure and staff, it is going to be necessary to develop the Cloud IT separately. The technologies, culture, Dev/Ops expectations, and services within the Cloud are too different from the ‘standard’ IT environment. The Cloud Group will be mandated to use Cloud technology to improve business processes and IT resource usage.
Failure is an option
Cloud platforms are complicated and the architecture design will be inclusive of best practices, and will be quite different than normal internal IT architectures. In migrating, moving or building greenfield Cloud systems, you will experience a learning curve, failure and especially around DevOps, a great deal of cultural adjustment.
Security Model
Once you have decided which type of Cloud deployment you are going to implement, and why; you need to layer in Security, and understand how Security works on your chosen Cloud platform [3rd party, AWS, Google, Azure]. Best practices around Security [IAM, Groups, ACLs, Bastion hosts, IP level listing etc], are mandatory.
Business Value from the Cloud
A main issue and concern is driving real Business Value from Cloud adoption. Anyone can spin up VMs and deploy a service and even forklift a million lines of code into a Cloud platform and allow remote access. This is not going to transform your business or add value.
If you really want drive business value, organizations need to change how applications are designed and how they function and most importantly, how they support a mutable business model. Applications must be Cloud-friendly and include inter-alia design choices around: session state, micro-service design, Rest, SOAP, multi-tenancy, security, fail-over, portability, de-coupled code, components, and mobility. CI and CD [continuous improvement, continuous delivery], and Agile programming cannot be done within the traditional IT silos. Discrete functionality being produced and deployed continuously means that your architecture must be Cloud friendly, including the code base.
Data Models
IT always comes back to the Data, and specifically the quality of the Data. You can’t move to the Cloud if your Data is bad, of unknown quality or poorly coded. You must ensure that the Data is understood, the data-logical-conceptual models exist and are valid, and that the code is documented. Data strategy and data modelling are the first steps in moving to the Cloud.