Application performance in the virtualized data centerEnd to end visibility is needed Tom Joyce, President & CEO, Akorri Highlights
|

According to the Aberdeen Group, lack of visibility into the entire transaction flow and inability to anticipate performance issues while conducting virtualization projects are the top obstacles for achieving the optimal level of application performance in virtualized environments. Managing virtualization performance has become top of mind as companies are now trying to do more with their virtualization environments including virtualizing production applications. Because of the complexity of these environments, performance management tools used for physical environments won’t work and virtual machine provisioning and availability monitoring isn’t good enough. Companies now need tools that provide visibility across virtualization layers and IT silos and can identify and troubleshoot application issues to ensure good application response time.
According to the Aberdeen Group, December 2008, the top obstacles for achieving the optimal level of application performance in virtualized environments are:
--Lack of visibility into entire transaction flow-55%
--Inability to anticipate performance issues while conducting virtualization projects-49%
--Inability to measure quality of end-user experience-46%
--Inability to manage SLAs around application performance-43%
Cross-Domain Visibility Is Required
Just like a car mechanic needs to “lift the hood” to optimize and tune an engine, IT managers need to see inside and through virtualization layers. Application performance relies heavily on underlying infrastructure. To optimize virtualized applications’ performance, it is necessary to be able to map logical data paths across each virtualized layer and IT domain down to the underlying shared physical resources, with applications, servers, and storage fully mapped together. Mapping an application’s logical data flow enables IT management to see which actual resources are being used by an application. An application’s performance in abstracted virtual IT layers can be analyzed by applying dynamic queuing models built over the observed application and infrastructure data. This results in an automatically generated performance map that provides visibility into the entire application flow.
Visibility Enables Predicting and Resolving Performance Issues
Having the entire virtual infrastructure in a performance map allows IT managers to quickly identify problems, such as contention points and bottlenecks, and minimize disruptions in the future. If end users are experiencing latency, it is critical to have the intelligence to understand the problem’s root cause. Often it’s the storage layer that presents the greatest management challenge. Storage plays a critical role in virtualization, as multiple virtual machines across different physical servers can share the same set of disks on a storage array. The result is often contention between virtual machines that cause performance issues within the virtual infrastructure, and ultimately, negatively impacts applications and end users. Good cross-domain performance models can immediately show where to find and how to remediate performance constraining bottlenecks, judge and produce optimizing recommendations for aligning resources to applications, and help plan future capacity needs.
Use Cross-Domain Tools to Ensure Application Performance in Virtualized Environments
To understand and resolve performance issues, IT managers need tools that analyze virtual and physical infrastructure performance giving them end-to-end visibility that shows the interdependencies between virtual and physical systems. With these tools, IT has the confidence to virtualize more business critical applications and push VM densities to get more out of the infrastructure while ensuring good application performance for end users.
Tom Joyce is President and CEO of Akorri, a virtual infrastructure management software company. He can be reached at . Or visit their web site at http://www.akorri.com.
Comments:









