As technologists and analysts for the virtualization and cloud spaces, we are always talking about various places within the IT stack. Actually, as we talked about within the article Technical Arc of Virtualization, we have noticed that many people are moving up the IT stack, forming new and more interesting substrates of IT. These substrates are used to simplify the actions one takes to deploy new and more interesting applications, while at the same time abstracting away the physical and virtual layers of the stack—in essence, forming new substrates on top of which to build.
The following infographic (click to expand) depicts the substrates of IT and where each belongs within physical, virtual, and cloud environments. Within the cloud, for example, the physical and hypervisors are abstracted away. Within a software-defined data center (SDDC), we even abstract away the virtual networking components, relying entirely upon software-defined networking and network function virtualization to deliver our needs. However, even as we move up to higher substrates, we cannot yet completely ignore the virtual machine. We are getting close with containers, where deployment of an application will abstract all below, with only the application mattering.
In the infographic we have shown where our sponsors (at the time of writing) fit within these substrates. Each companies products span physical, virtual and cloud environments so placement is not really important within each substrate.
The really interesting thing is how these substrates of IT have changed based on feedback from the monitoring and operational tools—the top substrate. Analytics has made automation king, with automated deployments, network creation, etc.—even deployment into clouds across the spectrum of the hybrid cloud. To make this possible, there is a growing substrate that comprises IT Management/Agents. In order to abstract everything below, the number of agents is increasing, the number of ways to deploy something is increasing, and the number of tools is increasing. The width of this substrate is growing considerably.
The IT Management/Agents substrate controls everything below it and quite a bit above it across all environments. Feedback into this layer triggers other actions that scale the environment. Managing this and the deployment layer is starting to get tougher and tougher. As more is put into it, we need better tools to manage everything. This IT substrate grows in conjunction with operations and the number of applications deployed.
We are also noticing that the Deployment substrate is growing, as more and more development concentrates Infrastructure as Code, Security as Code, Testing as Code, and code in general. New and interesting techniques are falling into this substrate, all designed to deploy the application while managing the software that creates the modern data center across all environments.
Yet, there is also a renewed interest in the physical layer with hyperconverged infrastructures—self-contained physical appliances that include all substrates below the SDDC substrate. This started with the VCE Vblock but has grown to include a number of purpose-built, hyperconverged infrastructure, appliances that provide all that is needed to abstract the physical: to provide a software-defined data center.
Many companies have products within each substrate, while others concentrate on just one substrate. The number of substrates and their importance is changing within IT, yet many substrates of IT still remain even if people move up the stack (of substrate layers). All technologists and architects need to determine where the substrates are within their organizations and whether they matter to current and future implementations.
The substrates of IT also define the new teams within IT, teams that work together to fully optimize, automate, and control the new IT. In this new IT, there are no silos, only in-KNOW-vation. The new teams work together to create highly scalable, automated deployments across physical, virtual, and cloud environments to meet current and future business needs on a quarter-hour basis, including show-back into costs and savings. This is the goal of the software-defined data center, yet we are still building there. Pretty soon, we will have all the tools we need to just write the code that meets our needs without having to worry about everything else, as everything else will just be code and abstracted away.
In the future, perhaps the only folks who will worry about everything below the SDDC substrate will just be service providers!
Share this Article:
Latest posts by Edward Haletky (see all)
- Common Product Security Questions - November 23, 2016
- Sorry Support: Not Getting My Data - November 18, 2016
- Moving to the Future: Strategies for Handling Data Scale - November 14, 2016