Blog

Advantages and Disadvantages of Cloud

January 30, 2018 | Ken Leoni

When C-level management and IT professionals clearly understand the potential advantages and disadvantages of cloud computing they can better develop a cohesive cloud strategy that both addresses the organization's technical requirements and is also economically sound.  

Business man under a cloud with arms open - isolated over white.jpegCloud computing is booming not only because of its technical and economic merits, but also because organizations are racing to adopt it out of fear of being left behind or at a competitive disadvantage.

The decision for cloud adoption can have a lasting impact on an organization.  Whether that impact is positive or negative will largely depend on the due diligence that is performed and what factors are incorporated into the decision. 

Here is a bit of guidance to help with the development of an informed cloud strategy.

 

Cloud Advantages:

No need to invest in a physical infrastructure – A cloud hosted infrastructure greatly reduces capital outlay as an organization is renting resources and software rather than  purchasing hardware and licenses.  The level of benefit depends on a number of factors:

  • If an organization is just starting out then the economics of cloud make quite a bit of sense as the upfront cost are limited to Opex.  In addition rapid expansions can be more easily handled, especially when there is limited cash flow. If for some reason there is contraction, unlike in the case of on-premises environments, the organization isn't saddled with an expensive physical IT infrastructure.

  • If hybrid cloud is part of the equation then it is good idea to pay special attention to what resources are allocated and where. Organizations that are under pressure to move to the cloud “because everyone else is doing it” can make costly decisions. For example, if workload(s) or application(s) are moved to cloud make sure there is a commensurate reduction in on-premises hardware and software licensing.
     
  • Cloud cost savings are more easily realized when the on-premises infrastructure is older and at capacity. If the data center is “holding its own” and the sunk cost have already been absorbed cloud may not make sense.

  • Volatile workloads, especially ones that are short-lived are ideal candidates for cloud. It is hugely expensive to have idle on-premises / private cloud resources at the ready in order to handle a spike in workload. It can be more cost effective to rent IT resources for periods of time (cloud) verses owning in perpetuity (on-premises).





Cloud platforms can be easier and quicker to administer – As the ecosystems tend to be well integrated, specializing in specific cloud technology allows IT to efficiently tailor and implement cloud services.  Cloud products and services for AWSAzure, and Google are numerous and growing.

  • Cloud deployments abstract the hardware for IT.  Not only can cloud administration be quicker, albeit with a learning curve in the beginning, but IT is assured that all components are compatible with each other.

  • However on-premises / private cloud  deployments require IT to maintain core competencies in a number of disparate technologies.  For example, IT needs to understand and integrate servers, storage, network, and any new hardware.

  • Another downside is vendor lock-in. Moving from one cloud platform to another can be painful and expensive, so you will want to be confident in your initial choice.



 

Cloud readily handles data locality –  If the nature of the business mandates relatively low latency or that the data be stored near where it is processed then cloud is an excellent option.

  • The business case is very compelling especially when it is necessary to deploy cloud resources in geographically dispersed locations. Leveraging a single set of IT resources to managed it all means no need to travel, no need to build out a data center, and no need to maintain IT personnel at remote locations.

  • As cloud adoption continues to expand, so do the number of hosting locations. For example -  AWS, Azure, and Google are all growing their worldwide data center presence.





Cloud is more resilient –  A cloud providers livelihood is contingent on them delivering a highly scalable and reliable infrastructure.


In addition to the resiliency of the hardware, we must also consider the software as well. Cloud services like Database-as-a-Service are resilient by design – any applications that are built using these services will also be inherently redundant.

 

Cloud can be more secure -   Security concerns should be rooted less on where the data is located and more on how it is secured.


Cloud can be more secure because not only can the data center itself be more physically secure, but also cloud services (when deployed based on best practices) are fundamentally secure.  

In addition, cloud providers can afford to invest in the expensive personnel and other resources required to ensure proper security. This is a boon to cloud customers who would otherwise have limited resources to dedicate to security.


 

Cloud Disadvantages:

Cloud can be quite expensive –  Cloud is touted as a cost saver, and while there this certainly a great deal of merit to cloud, depending on the configuration it may not necessarily result in any savings at all – in fact it can end up being quite a bit more expensive.

  • While it is easy to focus the spend for cloud on tangibles (i.e. server and storage) it is important not overlook software licensing as a significant part of the equation.

    Let’s consider something as simple as Windows based IaaS -  where the provider is bundling and charging monthly for the IT infrastructure and Windows licenses.

    The IaaS Windows licenses could well cost more than on-premises licenses because the subscription payments over time may exceed the cost of the perpetual license. If the licenses are instead purchased up front and implemented as part of a private cloud then a one and done purchase could prove to be the most cost effective choice.

    A large enterprise that is operating a long-term stable workload may  achieve better savings with an on-premises / private cloud deployment over outsourcing to a cloud provider.


  •  Improperly allocated workloads quickly escalate costs.

    It is critical to understand what the current workload looks like now as well   how the workload will grow. Plan too low and you may not be able to take advantage of the price breaks that come with cloud economies of scale. Plan too high and you’re either paying for unused resources or having to sell the surplus at a discount.

    Cloud Capacity Planning and performance reporting are absolute musts if an organization is to ensure that it is making the most efficient use of the resources.

  • While cloud adoption continues to increase, the available pool of talent to manage cloud still remains at a premium.  Cloud deployments require IT personnel with new and very different skill sets - attracting and retaining that expertise can be a challenge. Organizations need to understand that cloud implementations require IT departments to either import or nurture cloud competencies, that developing or bringing in cloud proficiency can be expensive, and that movement to cloud doesn’t necessarily translate to a reduction in IT personnel.
  • Deployments where cloud proficiency is lacking can be expensive endeavors as there is an inordinate amount of time spent managing cloud services and cloud resources may not be optimally configured.

 

Capacity Planning for Cloud Resources
Longitude Capacity Planner showing current VM workloads on a host




No access to the physical IT infrastructure –
While cloud providers have gone to great lengths to give users control over cloud services, access to the back end infrastructure is off limits. 

  • IT often wants a holistic picture of what is happening behind the scenes.

    Cloud providers often operate using a shared IT infrastructure model making it impractical to provide organizations access to the back end infrastructure. Some would say an environment where resources are shared is itself problematic, however the major cloud services providers have done an exemplary job of building an infrastructure to handle the demand.

  • Surrendering physical control can be a cause for concern for security and compliance reasons.

    Insider attacks are an issue whether the infrastructure is on-premises or cloud. However with cloud, IT is now relying on the providers’ internal policies and procedures to properly secure and segregate data, as well as control access to the physical infrastructure.

    Ultimately it may come down to an organization’s comfort level with having its customer sensitive information residing in the public cloud.




Cloud provider lock-in – The implementation of cloud services differ between providers.  For example, if we compare AWS and Azure and look at their serverless computing offerings;  AWS Lambda and Azure Functions - switching from one platform to the other is not a simple lift and shift.

  • Switching between cloud platforms becomes more challenging as organizations take advantage of the increasing number of cloud service offerings from a given provider.  Although the services are tightly integrated which allows for the most efficient use of IT and development resources - it comes at the cost of portability.  

  • There are those that advocate for a multi-cloud strategy whereby cloud services are employed from multiple cloud providers. The theory being that you are hedging your bets. The trade off is additional flexibility at the cost of a whole new level of complexity.

  • When committing to cloud service provider(s), organizations need to understand that this is a long-term relationship – that it is in fact a marriage of sorts - and that a divorce will likely be ugly and expensive.

 

Conclusion:

Cloud computing holds distinct economic and technical advantages in environments where workloads dramatically change as IT organizations can limit their expenditures with a pay-as-you-go payment model.  In addition cloud services allow organizations to develop and deploy highly scalable applications very efficiently.

However, migration to cloud is not a panacea. IT organization really need to do their homework as cloud costs can quickly escalate. Due diligence includes calculating the projected cloud costs over a 3 to 5-year period and comparing to the sunk and variable costs of on-premises / private cloud computing.

Calculating cost isn’t as simple as translating the current on-premises servers and storage to a lift and shift to cloud. Instead one needs to aggregate the on-premises resource utilization and properly translate and redistribute to the new cloud environment, the goal being to avoid over provisioning and unnecessary subscription costs.

Longitude Report showing VM Sprawl for EC2 Instances
Longitude reporting on over provisioning of cloud resources

 

Like most everything IT, there is no right or wrong answer.  Each environment has a unique set of requirements and challenges. As long as organizations can get past the hype and focus on capabilities and actual cost they will be able to make sound decisions.

 

Want to learn more?

Download our AWS vs Azure comparison matrix -  Amazon and Microsoft each offer cloud technology and services that are excellent in their own right. Which attributes and capabilities are a better fit for your organization.

 

Download the comparison:   AWS vs Azure 

Sign Up for the Blog

Heroix will never sell or redistribute your email address.