You can find many articles on cloud computing on the web — the why’s and wherefores, the dos and don’ts — but what motivates companies to start using public clouds for compute services? This blog explores the pros and cons of cloud computing as shared to us by our customers.
This is often cited as the primary reason for any move to the public cloud. The cost of on-premise IT is often difficult to define; not just the capital expenditure of hardware and software amortised over a number of years, but also the hard-to-define costs of repairs, replacement, updates, security, compliance, racking, lighting, cooling, networking, backups, disaster recovery arrangements, monitoring, Ops teams, etc., etc.
Public clouds allow a shift from a Capex to a mostly Opex model, and with them a transparency and predictability to the hidden costs.
On the other hand, the very transparency of public cloud costs is sometimes quoted as a reason not to adopt them. Public cloud costs are very apparent and clearly laid out in monthly bills – something that most CFO’s and Financial Controllers are not used to seeing or – more likely – cannot get for on-premise IT. This “sticker shock” is often quoted as a factor in the perception that public clouds are more expensive than on-premise IT.
For many organisations the costs of proof of compliance with the myriad security and privacy programmes such as ISO270001, ISO 27017, PCI, SOC I, SOC II, SOC II, HIPAA and others is simply far too expensive. Public clouds – especially the “big three” of AWS, Azure and the Google Cloud Platform – offer compliance with many of these schemes as part of their offering, amortising the costs of compliance into the rental rate of their computer services.
This assumes you only store data in the cloud – if you use a mixed model then these benefits can often be moot.
This is often overlooked, but running an on-premise data center necessarily means buying and installing hardware. This is yours for whatever period you plan to run it and so is the performance. To improve performance, you typically need to re-invest. Public clouds on the other hand offer continual improvements in performance through new compute service offerings, and taking advantage of this is typically no harder than moving a workload, often gaining performance at no extra cost.
But what happens when performance isn’t what you expect or think you are paying for? The root cause analysis around aspects like this can be daunting, and cloud compute performance can vary considerably across supposedly “equal” resources.
As your requirements change, public clouds allow you to trim your sail to suit your needs. As requirements for compute services grow (or reduce), public clouds allow you to simply increase or decrease your usage. No need to invest in new hardware or have unused hardware sitting around. Additionally, public clouds allow you to situate compute services in optimal geographic locations around the world.
This flexibility is all well and good, but the costs of getting there – migrating existing on-premise systems to the cloud – are often hard to predict and introduce degrees of risk that frequently halt migration projects.
Organisations are sometimes concerned about where (geographically) data is stored – be it for political or regulatory reasons. Public clouds allow easy selection of many geographic locations for both processing and storage of data, requiring no investment in private of co-located data centres around the world.
Of course, this does not help the “ownership” issue – who “owns” the hardware where the data is stored.
The nature of public cloud architectures makes it relatively easy to recover from most forms of system failure by simply moving to new compute resources at no extra cost. Long term storage is relatively cheap and distributed for reliability. There is no need to invest in DR systems simply to be available should they be needed.
On the other hand, if a public cloud provider suffers a total failure themselves (rare but not unheard of), cloud-based DR with the same cloud provider offers little and – depending on your needs and appetite for risk – some customers have decided to have DR in a second cloud providers cloud or even on-premise.
It is easy to see how the advantages of cloud computing easily outweigh the drawbacks. There is no doubt that businesses can reap huge benefits from cloud computing; decreased costs, reduced downtime, and less management effort are benefits that speak for themselves. However, with the many advantages, come some drawbacks as well.
With careful planning, precautions and the ongoing optimisation utilizing IT asset intelligence across the entire IT estate, the disadvantages of cloud computing can be minimized. Take time to understand both the pros and cons of cloud computing, so that you can get the most out of your business technology, whichever cloud provider you choose.
Learn more about how you can plan your migration to the cloud in our Smart Guide: Essential considerations for cloud migration planning and cost optimization.