![Industry Perspectives Industry Perspectives](https://eu-images.contentstack.com/v3/assets/blt8eb3cdfc1fce5194/blt79a1824cfad941c4/664ba129be90a086233eb8e0/Industry_Perspectives_DCK_light.png?width=700&auto=webp&quality=80&disable=upscale)
Insight and analysis on the data center space from industry thought leaders.
Are Your Enterprise Applications Ready for an OpenStack Neighborhood?
When running applications in a multi-tenant environment, you need assurance of performance and reliability. Keeping an eye on "noisy neighbors" can be important, according to Billy Cox of Intel.
May 21, 2014
Billy Cox, the general manager of Service Assurance Management for Intel.
The typical enterprise application grew up on its own block. By that I mean the application lived in its own environment served by a dedicated pool of bare-metal servers. For the administrator, there was no question the application would get the resources it needed to meet the appropriate service-level agreements (SLAs). That was all hard wired into the system.
Today, with the advent of cloud computing and the maturation of the supporting technologies, a big change is under way. Enterprise applications are now moving into multi-tenant OpenStack neighborhoods where they have to contend with potentially “noisy neighbors.”
A so-called “noisy neighbor” is a virtual machine that consumes more than its fair share of the shared resources within the system and therefore degrades the performance of the other VMs running on that same system. And this is where things get even trickier: When you move an application into a shared environment, you might encounter performance problems but have no way of knowing they are due to a noisy neighbor.
At Intel, we are addressing this problem with software, called Intel Data Center Manager: Service Assurance Administrator (Intel DCM: SAA). This software plugs into OpenStack to extend its functionality in a natural way. It uses OpenStack methods to determine the best placement for a particular virtual machine and the workload it runs.
Intel DCM: SAA has built-in capabilities that use Service Level Agreement (SLA) specifications as performance targets. The details on these capabilities are beyond the scope of this blog post. For now, let’s just cut to the chase: By assigning a specific SLA to an application, all stakeholders can have the confidence that—regardless of other activities in the neighborhood—the application will get the CPU resources it needs to meet the expected performance criteria.
Your workloads also want to live in a secure and trusted neighborhood. You need the assurance that the neighborhood is in fact the right neighborhood (the right hypervisor) and is running in compliance with required regulations (configured to meet requirements).
What application owners need are assurances about the performance and trust of workloads running in a multi-tenant environment. Service providers, in turn, gain the means to create new offerings with assured performance or to pack workloads more densely into existing server racks—with the confidence that they will meet their SLAs.
Looking ahead, Intel plans to add to the software with quality of service capabilities for storage and networking resources—all in the interest of maintaining good neighborly relations on OpenStack blocks.
For a closer look at the capabilities in the initial release of Intel DCM: SAA, watch this video and take a look under the hood. Let me know what you think on Twitter @IntelITCenter.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.
About the Author
You May Also Like