Making Sense of Cloud Computing
Most long-existing companies are facing the reality of needing to keep up with the fast-paced IT world and fighting competition from new start-ups taking advantage of all the bells and whistles of giant cloud services providers such as Google Cloud Services (GCP), Microsoft Azure and Amazon Web Services (AWS).
It is no longer a question of ‘if’ businesses should make use of the cloud as part of their technology arsenal, but ‘how to’. The first step is to understand the differences between services that are built for the cloud (cloud-native), as opposed to managed services (products simply hosted in the cloud).
Why is the cloud an almost de-facto standard choice for most start-ups?
Simple. It solves several blockers to start a business, such as needing a large capital upfront to purchase physical servers; hosting and cooling and securing the servers; getting a fast internet connection; hiring large teams of professionals to combine all technologies efficiently; potentially developing basic building block components to interface with 3rd party systems.
Starting cloud-less/on-premises would exponentially delay the initial phase until your business rewards start coming in. You would also be reinventing the wheel on many fronts to cover basics that are solved with cloud services which are cheap and available at the click of a button. This would put you at a substantial time and money competitive disadvantage.
What are the advantages of going ‘cloud’?
Going cloud gives you instant access to different tiers of modern technologies to cherry pick from anywhere in the globe at the click of a button. Moreover, there are generally no upfront costs, and you pay for these services by usage, typically hourly. This provides incredible flexibility to combine the right technologies without breaking your bank account and without specific commitments.
With access to this setup, the prerequisites to start a business have changed drastically. All you need is a handful of technical personnel who know cloud services well, a handful of developers and the rest of the traditional supporting business functions. The efforts would be concentrated on the actual solutions that will eventually deliver business value, and not on all the building blocks that would typically be blockers in an on-premises setup.
Is the cloud less expensive in the long run?
The short answer is no. The cloud typically is more expensive when used long term. However, you need to take into consideration a lot of factors. By utilising cloud services, you are not just getting a bunch of servers in return, but also all the surrounding infrastructure and services that provide the complete readily-available solution.
Included in the usage pricing, you typically also get:
– Electricity and cooling
– High speed internet connectivity
– Strong, secure locations
– Redundancy of components at all levels (Network, Storage, Servers, etc.)
– Software licenses (Operating System, Applications, etc.)
– Marketplace with built-in automation and templates options
– A level of support (based on the tier you pay for)
– Included future hardware upgrades (without having to setup again from scratch)
The on-demand pricing is ideal to start setting up cloud services and to experiment with the right configuration. It also provides flexibility to ‘burst’ your resources to momentarily handle peaks in demand, such as when there is a heavy promotion or campaign going on. The hourly usage pricing is also ideal for temporary setups used for throw-away prototyping for example.
A commitment (or reserved instances) option is also available for workloads that will persist 24/7 for over a year. This would typically present very good discounts of 20-40% and sometimes more, depending on the service and the length of the commitment.
What does ‘cloud-native’ actually mean?
There are a few main approaches that are used to migrate and/or host the cloud services:
Lift and Shift/Rehosting
Refers to migrating an already existing system from on-premises to the cloud without transforming the system first. To replicate an on-premises setup, the use of mostly Compute (traditional Virtual Machines) would be prevalent. This is called an Infrastructure-as-a-Service (IaaS) approach and it is the costliest way of utilising the cloud. Essentially, you would be just renting someone else’s computer and paying by the hour.
A better approach would be to reshape an application to make use of a Platform-as-a-Service (PaaS) or Software-as-a-Service (SaaS) approach to make use of already established services in the cloud that are readily available. This can scale up/down as needed and would be priced by usage of that specific service.
As the name suggests, a hybrid approach to cloud hosting would be using services from the cloud as well as your own on-premises services. Typically, this would be required as a phased approach to transition to fully cloud-hosted services, however a hybrid-cloud approach can also be used strategically for cost, performance, technology limitations, regulations, jurisdictions, and other limitations.
Why limit yourself to just one cloud provider? You can use more than one cloud provider strategically to access different technologies, geographical restrictions, backups, disaster recovery and cost reasons.
The matrix below shows the different ‘as-a-service’ approaches using the cloud shared responsibility model to indicate areas of responsibility:
When not to use the cloud?
The cloud is not an answer to all problems and situations. There are certain situations where the cloud can be a bad choice such as:
Some specific workloads require a combination of hardware not available on the cloud to achieve a certain level of performance.
Some laws require specific data or workloads to exist in specific locations and under full control of the company.
- Technology Limitations/Availability
Certain technology advancements might have not reached the cloud.
A certain combination of services can be prohibitively costly to ‘rent’ in the cloud compared to purchasing your own Infrastructure.
Lack of Knowledge
If there is a lack of understanding on how the cloud and its various services operate, then it can be complex or inefficient to use.
Certain policies might prohibit cloud use.
The Cloud for Northern Gas and Power
Here at NGP, the cloud has been embraced from the start of our technology division. With just a handful of employees in small teams of Developers, DevOps and Infrastructure, NGP was able to leverage cloud resources that would otherwise have required 5 times the team sizes and more upfront capital to achieve in such a short time.
With the right people and mindset onboard, NGP has built ClearVUE. Lite and ClearVUE. PRO, and an IoT Data Pipeline shaped to run cloud-natively using PaaS and SaaS for efficient use of resources and fast scalability.
By Jonathan Spiteri
Jonathan Spiteri is our Head of Infrastructure based in Malta. He holds a Masters degree in Information Technology and possesses multiple hands-on certifications on cloud, virtualisation, Windows and Linux. Jonathan’s key focuses are IT operations, development, problem-solving, large-scale IT-design and implementation focusing on innovation, automation, efficiency, team growth and mentoring.
He has 20 years of experience in multiple industries including government agencies, telecommunication, web application security, online casino gaming, online video streaming and banking. His experience has given him a clear perspective across the business, where he can guide the GPG teams towards the right choices, putting the group of businesses at a clear competitive advantage.