What is traditional computing?
Traditional computing is using physical hardware, (servers, storage, network equipment, etc.) running in a centralized location.
Usually these physical assets are located in a data center, either on premises or in a dedicated remote location. The bottom line is that the business is accountable for the physical hardware to handle data storage, network communications, application hosting, and security.
Advantages:
- End to end control – If the business wants to expand connectivity or re-swizzle the server configuration, they are completely free to do so.
- Predictable cost – Once everything has been provisioned there is little fluctuation in the cost, assuming the utilization requirements don’t change.
- Performance – Dedicated physical hardware, which the business isn’t sharing with other businesses, can often be more performant. (As an aside, it is possible to rent shared virtual space on a server. In this case performance is usually throttled and the benefit degraded.)
Disadvantages:
- Maintenance and support – The business is responsible for all of the upkeep of the hardware. From replacing defective hard drives to patching the operating system, the business is on the hook. As the saying goes – “With great power comes great responsibility”.
- Limited scalability – As traffic and consumption of resources grows, so does the need for more/better hardware. This has other costs associated with it more than just the hardware itself. Things like:
- Engineering effort to incorporate the new hardware.
- Downtime while upgrading.
- Decreased traffic, leaving the business with too much hardware to utilize.
Limited ability to scale up and down is a major disadvantage of traditional computing.
- Limited reliability – Physical hardware breaks down and becomes obsolete. Even the best designed systems will suffer outages. Owning the hardware is a risk that needs to be assessed; downtime is expensive.
What is cloud computing?
Cloud computing is utilizing compute and data resources in a decentralized model.
The cloud provider gives the business access to things like data storage, compute resources, and network gateways in the form of commodity services. In addition, the cloud provider also gives access to specialized services where the hardware/infrastructure complexity is abstracted away. A summary of these services is outside the scope of this article but, suffice to say, these can be extremely cost effective.
Advantages:
- High reliability – As cloud computing is decentralized, oftentimes redundancy is handled for the business. Services are “always available” with numbers touted like 99.99% uptime. That number is effectively impossible with traditional computing.
- Enhanced scalability – Scaling up or down is made simple in the cloud; allocating and removing resources is a core functionality that most cloud providers offer. This means that you are only paying for what you use.
- Large offering – Cloud providers like AWS, Azure, and Google Cloud have an extremely wide variety of offerings. With these services, businesses have a huge amount of power to deliver digital offerings with incredible convenience and speed.
Disadvantages:
- Fluctuating costs – Costs can easily get out-of-control in the cloud. Although most providers provide a tool to calculate usage, real costs can still be vague and hard to quantify. For example if traffic were to spike in an unexpected way, the business may be subject to an unexpected cost increase.
- Intermittent access – The cloud is generally extremely reliable, access to it on the other hand can be a factor. Access to cloud services are subject to network connection and availability. Dedicated network connections are available but come at a very high cost.
- High learning curve – With the breadth of service offerings available, this also comes with a relatively high learning curve. Entire education paths with multiple certifications are available in dozens of specialities. Designing systems to run efficiently in the cloud requires specific knowledge and careful planning.
When to use traditional computing?
Using traditional computing can still be very cost effective assuming a couple things.
- The business is aware of the entire scope of resource consumption (things like storage and traffic) and there isn’t much fluctuation.
- Speed of delivery is a secondary or tertiary consideration. It does take time to provision hardware.
Scenarios best suited to traditional computing:
- Requirements for data privacy are such that the data needs to be secured in a physical location.
- Direct dedicated network access to compute resources is required.
- The business already owns the hardware and has available resources. (A cost analysis may be needed to understand the cost of leveraging existing resources vs utilizing the cloud)
- Diversifying a business’ technology portfolio. Utilizing physical hardware can be a way for risk averse companies to not be held to one methodology.
When to use the cloud?
In everything else.
In all seriousness the cloud is extremely flexible and capable and there’s almost no limit to what it can be used for. Assuming that a digital product was designed to be cost effective in the cloud, there are very few reasons why a business would avoid it.
Summary
The observant may have noticed that a lot of the disadvantages of traditional computing line up with the advantages of cloud computing and vice versa. This is somewhat intentional to show that these are opposing methodologies. Anymore, the question isn’t: “how to make the cloud duplicate what traditional computing can do?”, rather it’s “considering the requirements, which tool do we reach for?”.
This, obviously, isn’t an exhaustive list of cloud computing vs traditional computing but it does set the stage for making informed decisions. Hopefully this article begins to help you decide which tool to reach for. Thanks for reading.
If you would like to know more about out how I can help your organization in its cloud journey send me an email on the contact page.
