How does Distributed Power Management help server consolidation and save power at the same time?
Distributed Power Management (DPM) is all about getting more out of a good thing. Most IT professionals understand how virtualization allows multiple workloads to run on the same physical server. In this way, one server is doing the work of many -- resulting in fewer physical servers to buy, power and cool. But common virtualization usually provisions computing resources -- processors, memory, I/O and so on -- to each VM on a static basis. To change those resources requires manual intervention from an IT administrator.
While resource allocation is usually static, the actual resource use is not. Resources are allocated based on maximum VM demands, but lightly used or idle workloads may require far fewer resources. For example, enterprise workloads' need for resources may drop significantly overnight or during weekends when there are few users or computing tasks, such as Web page requests. The effective result is that idle virtualized workloads can still waste resources.
Dynamic resource allocation tools such as VMware's Distributed Resource Scheduler (DRS) alleviate the need for manual intervention, allowing resource allocation and workload placement to be adjusted automatically based on activity levels defined by rules. The idea is that DRS can see when VMs are idle, and move even more VMs onto fewer servers. This increases the consolidation level and maximizes resource utilization. As activity levels and corresponding resource demands increase, DRS can redistribute VMs back onto more servers while reallocating more computing resources.
Unfortunately, the automated behavior of DRS -- and similar tools -- has limited benefits by itself because physical servers that have been emptied of their VMs still continue to run and consume energy. Most idle server hardware will still demand at least 60% of its maximum energy usage.
To get strong benefits from such dynamic consolidation, it would be ideal to power down the empty servers, then power them back up again before restoring the VMs to their original state. This is the role of DPM, which is an optional feature of DRS.
Dig Deeper on DRS and DPM
Related Q&A from Stephen J. Bigelow
While the Windows Admin Center is one way to manage the Azure Stack HCI platform, you can also use traditional, battle-tested tools. Continue Reading
There are many tools available on the AWS Marketplace for QA testing, making it difficult to determine where to begin. What should an enterprise look... Continue Reading
Hyper-converged infrastructure that runs on Windows Server is not a new concept, but Microsoft's Azure Stack HCI program has one big difference from ... Continue Reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.