The story of the “as a service” sisters — SaaS PaaS IaaS

First, there was the data center and time shared compute engines.  Then there was distributed computing to the desktop.

IT as a Service
IT as a Service

  Today, discussions of cloud computing is all the rage. 

Be it Microsoft’s Office365, Amazon Web Services (AWS), Google Cloud Platform, or one of the many on site offerings, Cloud Computing is here to stay.   This article will begin by exploring some of the many cloud computing advantages and disadvantages (for there are some!).  Next, we’ll define the three primary “as a service” solutions.  Finally, we’ll apply Cloud Computing architecture and describe how “the cloud” can be used by real live businesses. 

Advantages of Cloud Compute models

Cloud computing
Cloud computing
  1. On-demand.  In Cloud Compute models, the services are “on demand”.  This means that instead of having to rent a physical location, apply for permits, purchase physical servers, standing up those servers in a physical data center, and hiring engineers and staff to run the data center, the Customer can focus on speed to market and stand up the cloud on demand.  This allows the equivalent of “leasing” equipment, instead of forcing large capital outlays.
  2. Rapid elasticity.  In Cloud Compute models, services can be expanded rapidly, and disposed of just as rapidly.  This reduces the concerns for oversizing or undersizing equipment purchases.
  3. Business Continuity Planning and disaster recovery.  Cloud compute offers location abstraction, where the Customer does not have need to control the geographic deployment area.  That said, if properly deployed, Cloud Computing models provide most of the computing infrastructure required to solve business continuity (BCP) and disaster recovery — all built into the deployment. That is, multiple geographically distributed compute solutions can be deployed, all without standing up independent physical locations.  Although this does not solve the entire Business Continuity plan (click here for a more comprehensive discussion of BCP), it goes a long way in the right direction.
  4. Security.  The Host company provides the physical security to the servers and datacenter.  Depending on the solution, the Customer is responsible for various levels of data security.
  5. Improved mobility.  All forms of cloud computing offer improved mobility for the workforce by centralizing the compute stack into a remote addressable solution.  There is no longer a need to create and protect a DMZ – if your employees have an internet connection, they’ll have access to the CSP.  

Disadvantages of Cloud Compute models

Sorry no Internet
Sorry no Internet

The cloud compute model is highly effective, and there are many reasonable advantages of moving to a Cloud Service Provider (CSP).   That said, there are disadvantages to any solution, and CSP is no different.  As with any solution, it is important to consider the CSP risks before fully embracing the architecture. Here we’ll explore some of the disadvantages.

  1. CSP Outages.  Unfortunately, like all cloud stacks, cloud providers also suffer outages.  If an outage does occur, the Customer may feel helpless in relying on the CSP in bringing the system back online.  That said, Outage risks can be overcome by building multiple cloud stacks with multiple CSP’s providing distributed geographic deployment. 
  2. Network outages.  Network outages do and will occur.  In a purely on site solution, Internet Service Provider (ISP) failures do not impact the business.  However, in a cloud solution, the ISP is a primary point of failure.  To manage these risks, multiple ISPs can be employed.
  3. Security.  While CSP’s offer tremendous Security value, there is a risk that policies are not followed.  Depending on the type of business you are running, some of those risks can be transferred by way of contractual language such as Business Associates Agreements in a healthcare environment.  
  4. Vendor lock in.  Vendor concerns exist with shrink wrapped software, and even more concerns exist for cloud services.  The customer must always be cognizant of vendor lock in risks.  For example, the customer should have mitigation plans in place if a vendor goes out of business, or if a contract ends, or if the contract becomes unaffordable.  These plans should be tested on a regular basis to confirm that all data is recoverable and the business is able to continue unabated.  

Types of “As a Service” solutions

Types of cloud
Types of cloud

Although there are many different marketing descriptions, there are three primary “As-A-Service” cloud compute service models.  Here are the three models.

  1. Infrastructure as a Service (IaaS).  Infrastructure as a Service is the most basic “as a service” model.  IaaS is a solution where the customer is provided the ability to provision processing, storage, networking, and other basic computing components.   The consumer does not control the underlying cloud infrastructure, but does control the operating system and applications.  The Hosting company controls the data center including physical access to the infrastructure, heating and cooling, insurance, and other infrastructure costs. 
  2. Platform as a Service (PaaS).  Platform as a Service is a solution where the Customer is controlling the platform from the point of view of the Operating System.  In PaaS solutions, the Hosting company provides a Platform Deployment Template (or a selection of templates).   For example, a PaaS hosting company may provide templates for Windows 10 or Linux with a specific number of CPUs, specific amount of RAM, and specific hard drive capacity.  The Customer has full control over, and full responsibility, for configuring the Operating System and any associated applications.
  3. Software as a Service (SaaS). Software as a Service is the most controlled “As a Service” solution.  In this environment, the Customer is purchasing access to a software package that is hosted on the Host company’s platform and infrastructure.  Typically, the SaaS solution is accessed via a web interface, or deployed application.  Configuration of the SaaS is limited to configurations within the software itself, not associated with the Platform (the operating system) nor the Infrastructure (for example the number of CPUs).  If there are speed considerations that are purchased, the speed considerations are related to artifacts such as “Transactions per Second” instead of the number of CPUs.

Real life examples

Example
Example

We’ve covered Advantages and Disadvantages of Cloud Service Providers.  Now lets explore where these CSP solutions can be beneficial.

  1. Say, for example, you own a retail store.  You regularly serve 100 customers per day, but on Black Friday your customer base expands 10 fold to 1000 customers.  Through a CSP model, you can rapidly expand the services to handle these additional customers for those days, then tear down the services after the rush.  There is no need to purchase hardware and deploy it in your own data center, just lease the engines to accommodate the surge.
  2. If a business is wrestling with the risks and unknowns of building out a full data center, it may be more reasonable to stand up cloud services as are necessary.  In this way, the business owners can focus on the business instead of managing a data center and the staff to maintain it.
  3. In a SaaS environment, a business does not have to be concerned with regular software updates.  Instead, the CSP host will maintain the SaaS environment, and the business can focus on the business needs.  Security risks are also reduced since the most recent software package is regularly deployed.
  4. If a business experiences a recession or other cut backs, the cloud expenses can quickly be reduced.  Due to rapid elasticity, the business is not at risk of purchasing and maintaining large unused data centers.

Where to go from here

What's next
What’s next

As with any solution, it is a good idea to outline the specific benefits and concerns that you will have as you explore cloud computing.  As a recommendation, I’d say jump to Cloud earlier rather than later, but be confident that the risk plans are defined and managed appropriately for your business.

Reach out to me with any specific questions.  As always, let’s be careful out there! 

 

 

Key acronyms and technologies

  1. AWS – Amazon Web Services
  2. CSP – Cloud Service Provider
  3. ISP – Internet Service Provider
  4. SaaS – Software as a Service
  5. PaaS – Platform as a Service
  6. IaaS – Infrastructure as a Service

References

  1. The NIST Definition of Cloud Computing“,  http://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-145.pdf
  2. “Advantages and Disadvantages of Cloud Computing”,  http://www.levelcloud.net/why-levelcloud/cloud-education-center/advantages-and-disadvantages-of-cloud-computing/
  3. “Google Cloud Platform”, https://cloud.google.com/products/
  4. “Benefits of cloud computing”, https://www.ibm.com/cloud-computing/learn-more/benefits-of-cloud-computing/
  5. “11 Advantages of Cloud Computing and How Your Business Can Benefit From Them”, https://www.skyhighnetworks.com/cloud-security-blog/11-advantages-of-cloud-computing-and-how-your-business-can-benefit-from-them/
  6. “Cloud Computing and Is it Really All That Beneficial?”,  https://www.lifewire.com/cloud-computing-explained-2373125
  7. “Why Move To The Cloud? 10 Benefits Of Cloud Computing”,  https://www.salesforce.com/uk/blog/2015/11/why-move-to-the-cloud-10-benefits-of-cloud-computing.html