This post discuss the technical pros and cons of on-premise infrastructure and cloud infrastructure.
What is cloud computing?
Cloud computing is on-demand delivery of IT resources and applications via the Internet. In an on-premise infrastructure, the applications are hosted on your local server environment. In the cloud, the applications are hosted on cloud service provider’s data center. The three primary cloud service providers are AWS by Amazon, GCP by Google, and Azure by Microsoft.
Features of on-premise infrastructure
Systems are designed to handle peaks i.e. the highest expected load on CPU, memory, storage, networking, and power. If the system is overloaded, the business will frustrate and lose customers. If the system is under-utilized, the business is wasting excess capacity.
The business is responsible for power, power backups, physical security, infrastructure upgrades, and staffing their office with qualified individuals to perform these tasks.
Key distinguishing features of the cloud
Cloud service provider is responsible for:
- access to high speed networking
- power and power backup
- access to unlimited CPU, memory, and storage on-demand
- business has the ability to automatically scale up or down as your requirements change
- business only pays for what is uses and does not have to procure excess capacity (CPU, memory, storage)
- Cloud service provider is responsible for hardware upgrades and patches
- Cloud service provider is responsible for physical security of the data center
Cloud is Agile
Businesses can invest in large on-premise infrastructure and later realize that they are consistently under utilizing and losing money. With the cloud, business can start with small infrastructure and scale up as their business grows.
Three main factors that influence agility:
- increasing speed
- ease of experimentation
- cultivating a culture of innovation
With cloud, the business can rapidly scale its infrastructure up or down. It can rapidly setup a new environment to experiment with different types of servers, storage, and configuration and discard what they don’t like. This allows the technical staff to try different solutions rapidly and innovate at a low cost.
AWS provides CloudFormation templates, which are essentially instructions to spin up a server. By using these template, the business can spin up identical environments for dev, testing, and production within minutes.
Cloud reduces security risks
Best in the industry experts work for AWS, GCP, and Azure. By using the cloud, your business delegates many tasks to them such as hardware upgrade, software patching, and responding to certain security threats.
Cloud is scalable
Scalable means the business can resize its resources as necessary. Suppose, the business is running a sports website. The traffic will increase many-fold during the peak season and drop dramatically off-season. The business scale up (procure more CPU, memory, and / or storage) when needed and then scale down (release CPU, memory, storage) when the excess capacity is no longer needed. Spinning up new servers takes minutes.
Cloud is elastic
Elasticity is the ability to scale up or down resources rapidly. In the cloud, a business can:
- rapidly deploy new apps
- rapidly scale up or down as needed
- instantly shutdown resources no longer required
Cloud offers low latency. Low latency means that the requested content is delivered rapidly. Cloud data centers are located close to major Internet fiber backbones. AWS, GCP, and Azure have data centers in key locations across the globe. Furthermore, they allow businesses to offer cached content from various locations. This way, they are able to offer low latency content delivery.
Cloud is reliable
AWS, GCP, and Azure are designed to handle service failures. There are numerous data centers across the globe and you data is replicate across multiple data centers. This way if one data center suffers from power failure, connectivity interruption, and another disaster, other data centers can entire continuity of service.
Data ownership in the cloud
In AWS, GCP, and Azure cloud, the clients retains full ownership and control over their data. They have access to real-time infrastructure monitoring.