Hi Bryan,
The typical benefits of cloud hosting most talked about are:
- scalability
- Pay-as-you-Go fee
- Business Continuity
A key opportunity, which is often underestimated, is:
Infrastructure as Code / Automation
Cloud instances should be provisioned and managed programmatically, via APIs, without any manual intervention. Code and configuration are written, Unit Tested via Continuous Integration, and deployed via Continuous Delivery. That code is ultimately responsible for the provisioning, updating, scaling, and de-provisioning of infrastructure.
You can appreciate why such a paradigm shift in managing infrastructures is often under-estimated; it is hard! But totally worth the effort! Traditional tech operations wouldn't know where to start. They often end-up setting up and managing cloud infrastructures in the same way they used to manage dedicated infrastructures, which is pointless.
Here are few example of why Infrastructure as Code / Automation is important:
- Set-up speed: an enterprise-grade and fully scalable infrastructure can be set-up in 30 minutes or less;
- Standardisation: the above set-up can be repeated 100 times, and every time the outcome is exactly the same;
- Quality: quality is repeatable and built-in;
- Evolution: unlike traditional infrastructures as we know it, cloud infrastructures should be able to evolve, even on a daily basis with no downtime;
Here are few examples of the benefits:
- Scalability: by automating the provisioning, configuration and application deployment it is easy and reliable to scale. It will be possible to adopt Scheduled Scaling, Pro-active Scaling and Re-active/Real-Time Scaling. On this topic there is nice video of from AWS re:Invent, "Running and Scaling Magento on AWS" (you can start from minute 35 if you are short of time);
- Security: with Infrastructure as Code you define and enforce the desired state of the infrastructure and what you want it to do. That fits very well with managing Security seriously. With less manual intervention needed by both developers and operations, there is further opportunity to do platform hardening;
- Disaster Recovery: Cloud providers, such as AWS, provide regions, each with multiple zones. That is, you can provision instances across different data-centres, within that geographical region. But let say an entire region is taken down (e.g. Virginia)! Assuming that you have a DB replica on a different region (e.g. London), which is feasible, with the automation it should be possible to re-create from scratch the whole infrastructure on a different region and be up and running within a few hours.
- Cost: obviously on of the benefits of scalability is reduced cost. Furthermore, pre-production environments usually don't need to run 24/7; hence, they can be scheduled to be up only during business hours, e.g. 8am to 8pm Monday to Friday. The ability to automatically schedule the provisioning and de-provisioning of pre-production environments can save over 50% on infrastructure cost;
- Agility: software development will greatly benefit from such automation! For instance, think of the level of validation provided when testing new features on a pre-production (scalable) environment that is almost identical to live. The ability to automatically and reliably deploy code directly from Git or via a CI tool any time you want.
And we have just scratched the benefits of Cloud Infrastructure!
As you can imagine, smart technical operation people can spend less time on "plumbing and wiring" and more time creating and delivering value to developers and businesses they support. This is where the "Devops Way" really comes in and can make a huge difference!
I hope this helps re-framing how Cloud Infrastructure / Hosting can really help and shift mind-set from hosting as a commodity (GB, MHz, SSD, Redis etc...) to Technical Operations as a serious competitive advantage.
Best,
Osvaldo
CEO of Elastera
