A recent study by Uptime Institute surveyed 250 C-level executives and data center/ IT managers at enterprises globally and found that 51% of respondents noted virtualization as the technology with the highest impact on data center demand.
As the demand to run more applications grows, so too does the demand on data center power and cooling resources. Virtualization technology enables data center operators to harness more computing power with fewer physical servers. Managing this dynamic workload, however, can prove difficult, especially with respect to a data center’s power and cooling system.
6SigmaRoom, the industry’s leading data center CFD simulation software, can model virtualization’s dynamic workload, as well as its impact on a data center’s power and cooling system. 6SigmaRoom tests “What-If” scenarios, so you can verify that your data center can effectively cool IT based on traffic needs.
What is Virtualization?
Virtualization “is the process of creating a software-based, or virtual, representation of something” (VMware 2019). Specifically, server virtualization enables multiple virtual machines—made up of an application and an operating system (OS)—to run on a single physical server (VMware 2019). By splitting the workload in this way, IT can run at higher (or even full!) capacity. For a more comprehensive look at virtualization, check out VMware’s video here.
Virtualization and the Digital Twin
The process of moving workloads from a server or a group of servers to another set of servers creates a dynamic, time-dependent workload. The continual movement of workload makes management tricky because cooling and power units need to be able to react accordingly.
With the digital twin, you can simulate the moving workload and its impact on the data center’s power and cooling systems. The data collected from the different simulations not only can be used for evaluation, but can also be fed into your data center AI system. The trained AI system can then respond to the dynamic workload accordingly, i.e. if the workload acts in X way, then your power and cooling systems will react in a predetermined and tested way, as outlined in 6SigmaRoom’s sandbox environment.
Specifically, within 6SigmaRoom the user can alter the load (i.e. where, when and how much) to test varying configurations and cooling system responses across time (as pictured below in Figures 1 - 4).
Figure 1. Initial power distribution per cabinet
Figure 2. The initial inlet temperature to cabinets, the temperature distribution in the room and the cooling performance of each cooling unit
Figure 3. How the power has moved to the different locations within the room
Figure 4. The corresponding inlet temperature to the cabinets, the temperature distribution in the room and the cooling performance of the cooling units
The above results indicate that the inlet temperature to the IT changes in a non-negligible manner. Visualizing and mitigating these types of variabilities allows you to be confident in your data center’s performance.
Use 6SigmaRoom to Manage Virtualization’s Dynamic Workload
Implementing server virtualization will lead to a more efficient use of your equipment; however, ensuring that your power and cooling systems can respond accordingly to the rising demands that come with virtualization can be tricky. Testing whether your data center/ IT and cooling system can successfully react to dynamic workloads is a necessary first step in managing these demands in a live environment. To learn more about 6SigmaRoom’s data center management capabilities, consider reading our blogs that answer the following questions: How effective is your cooling system? Does your business have the capacity to change?
Blog written by: Akhil Docca, Director of Marketing & Danielle Gibson, Technical Marketing Writer
Other Recent Posts
The Digital Twin for Hyperscale Data Centers
In a world where artificial intelligence and machine learning are the new frontiers, cloud applicat…
Why the pandemic has increased the importance of thermal design
The pandemic has moved most organizations further down the spectrum of risk aversion, meaning that…
22 May, 2019