Benefits of the Public Cloud

November 17, 2016

By Seth Patterson, Flexential Executive Director of Client Architecture Cloud computing is ideal for businesses looking for scalability, instant provisioning, virtualized resources and the ability to expand the server base quickly. Within the three cloud service models (IaaS, SaaS and PaaS), there are basically two types of IaaS clouds: public clouds and private clouds. This blog explores the differences between public cloud types and when to use each to maximize results. My next blog post will look at private cloud types. Before any cloud implementation, clients should perform an application assessment to:

  1. Understand application state and dependencies for cloud design options
  2. Document the processing requirements of each application (or application class)
  3. Identify security and compliance posture and needs for these workloads

What is a “Public Cloud?” There are many different types of “public” clouds in the market. Most people are focused on what the major hyper-providers can offer, like Amazon Web Services or Microsoft Azure, as the top examples. There are also public clouds that are more like “community” clouds where providers offer a shared cloud environment with secure multi-tenancy, but in a more managed environment with many services offered standard and optional. The true value of the hyper-provider cloud model is that there is an extensive and constantly evolving product catalog that allows a company to add many technical service layers into their applications or environments. It is designed at such a scale that the bursting of workloads for minutes or hours can be easily done as there is ample capacity for that flexibility. Consider a hyper-provider public cloud solution for these scenarios:

  • Web servers or simple websites
  • Development or sandbox engagements
  • The utility model allows a client to spin up a virtual machine for a limited amount of time and tear it down without a contract – a true “pay as you go” model
  • IT shops that want to self-manage their instance
  • Applications that reach peak loads periodically and then retract
  • Bursting strategies that necessitate a relatively simple method of ramping up and down environments
  • Deploying new applications/development that can take advantage of the services and micro services for faster code deployment

Consider an alternative public/community cloud in some of these scenarios:  

  • Mission-critical applications with high-availability and compliance or regulatory requirements 
  • Applications that require compliant solutions or data sovereignty
  • Want or need for dedicated hardware, storage, or network for performance or other reasons
  • Possibly secured databases connected to burst strategies in the hyper cloud
  • Managed services – need for “high touch” services for operational functions
  • Proprietary algorithms that run on specialized hardware
  • Back-office ERP or other sensitive and/or audited systems  The chart below illustrates typical application workloads and where they are often heading:  

Typical application workloads and where they are often heading