Introducing the Application Control Module
At VMTurbo our stated mission is to “Assure application performance while utilizing the environment as efficiently as possible”. In our latest release (5.1) we have taken a great leap forward in the pursuit of our mission by adding significant capabilities at the application layer. In particular we’ve added support for Java Application Servers – specifically IBM Websphere, Oracle Weblogic, Red Hat Jboss, and Apache Tomcat.
The first thing that we set out to do was to extend our ability to understand application demand. In application servers demand is best characterized by transaction throughput and transaction response time. We now collect these metrics from each supported application server and allow our users to understand how their applications are performing relative to their desired service levels.
With our application server support we allow our customers to automatically scale their applications based on VMTurbo’s comprehensive analytics.
For vertically scaling applications we will reconfigure the application server to the optimal settings for the most critical resources including heap and threads. If our analytics determine that the application server is under-provisioned or over-provisioned we will recommend the optimal allocation and automate the change if directed.
For horizontally scaling applications we will provide recommendations to provision or decommission new application server instances to match the throughput of the cluster to the incoming demand.
Holistic Resource Management
As always all of our recommendations are made in a holistic manner. This means, for example, that we will not recommend an increase of the heap unless there are sufficient underlying resources (virtual and physical memory) to support it. And if more heap is required we will first take the other actions necessary (for example, moving VMs to free up sufficient physical memory on the host) to make sure that the change to the heap allocation does not adversely impact performance (for example, increasing the heap allocation on a host with highly utilized memory will introduce undesirable swapping).
Likewise, if we are recommending that additional threads be allocated we will ensure that sufficient CPU resources exist to process the increased transaction throughput.
Using Application Demand to Best Control Infrastructure Supply
We believe that the only way to assure application performance is to control infrastructure supply in the context of understanding application demand. With the release of our Application Control Module and the support for Application Servers we believe that we have taken a great leap forward. But we are far from finished. Watch this space for more exciting announcements in the future for deeper and broader application support.
And if you want to start assuring application performance in your data center start your free 30-day trial now.