Forecasting Cloud Coverage: Public, Private and Hybrid Clouds Defined
Jon Downey, Woolpert chief innovation officer, kicked off our Geospatial Cloud series in January by touching on a number of areas that our clients are working through during their cloud journeys. This month, I want to talk about the decisions around private, public and hybrid clouds.
Let me begin by sharing an opinion: Very rarely do organizations go with a purely private or purely public cloud, especially in the context of geospatial data. I’ve seen 100% cloud-native applications much more often in young organizations that were cloud native from Day One, and far more often than not, organizations with 10-year or longer histories find that hybrid cloud solutions best meet their needs.
How does this relate to your geospatial cloud journey? Let’s dig into some definitions, move onto a practical story, then explore how to put the information to use as you select the right cloud solution for your organization.
So, what is the cloud, anyway? There is more than a decade of thought and writing on this topic, and indeed, the very definition of “cloud” has changed considerably over that time. Consider this:
“Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. This cloud model is composed of five essential characteristics, three service models and four deployment models.”
That description is from the National Institute of Standards and Technology (NIST), which goes on to describe those characteristics, service models and deployment models. A recent call for something more reflective of the times concludes with this excerpt from “What is ‘Cloud’? It is Time to Update the NIST Definition?” from Christine Miyachi:
“Perhaps the authors at NIST will update their model as definitions from standards organizations are more widely vetted and accepted. For now, IaaS/PaaS/SaaS remained firmly perched as the best simple definitions and they are sticking because they are still largely accurate. Although newer models better portray what is happening in cloud, the simple explanation of SaaS, PaaS and IaaS helps people get started in the complex world of cloud.”
Because NIST’s 2011 definition still works well for our purposes, I encourage you to read it in its entirety. The full paper is only three pages long but has valuable information and largely rings true in 2020. Here are some definitions that also will help:
Public clouds are those from which multiple tenants rent computing capabilities from a cloud vendor. Google Cloud Platform (GCP), Microsoft Azure and Amazon Web Services (AWS) are the three of the most well-known public cloud providers for U.S.-centric IT audiences. Woolpert’s STREAM:RASTER product is hosted on Google’s public cloud.
Private clouds live entirely behind corporate firewalls or trust boundaries. A private cloud does not necessarily include everything in a data center or any other individual location. It is used solely for the use of one organization, while sharing all the NIST-defined characteristics of a cloud computing platform. Woolpert’s SmartView Connect product is hosted on Woolpert’s private cloud.
Hybrid clouds mix elements of both public and private clouds.
Selecting a cloud (or combination of clouds) model for your organization involves four key actions:
- Begin with a public cloud before moving to a hybrid.
- Flatten the difference between the cloud options.
- Base your delivery selection on XaaS.
- Prepare to replatform.
For example, at Woolpert we have taken these actions along our own journey to a hybrid cloud solution. Although SmartView Connect is hosted on our private cloud, after the introduction of STREAM:RASTER, SmartView Connect was modified to be able to work directly with STREAM:RASTER as a data source. For more on that story and advice about making the best decision for your organization, check our recent guide: Private, Public, Hybrid: Picking the Right Cloud.
Geospatial Cloud Series 2020
September: Remote Sensing Data at Home in the Cloud
October-November: With Great Power of the Public Cloud Comes Even Greater Responsibility