Today's business and technology executives are focused on increasing the business value of their IT assets. Businesses are experiencing rapid growth in data volumes and technology options. This is resulting in escalating hardware and software licensing costs.

New technology layers and their associated abstractions are emerging at an accelerated pace. Each aim to increase the speed of implementation, deployment and/or lower maintenance costs. One negative aspect of this is the increasing complexity of the overall computing environment. Coupled with this is that information technology organizations are typically setup with application teams aligned along business lines and administration groups specializing on specific components within the technology stack (e.g. database, storage, operating system, network). Many of these have subspecialties that are further subdivided into other teams. These factors have combined to make it very difficult to unify the multitudes of professionals with different backgrounds and allow them to effectively communicate during critical situations or compressed timeframes.

The team at Aeysis are specialists in a broad range of technologies that take a holistic view to optimize and profile a full application environment without regard to the source of the pathology. We are as unobtrusive to your staff as possible and can find the majority of the information we need simply by observing a running system.

We have engaged with hundreds of customers - most in the Fortune 500 - where our primary goal upon arrival is to improve application performance, stability and scalability. With the methodologies we have created, we are often afforded the opportunity to identify potential problem areas that exceed the scope of the original statement of work. In those cases, we will provide detailed descriptions of the anomalies and proposed remediation techniques without requiring a change order. Throughout our engagements, we have provided thousands of scripts in many languages that companies still use today to evaluate their environment and ensure against performance or stability regression.

In addition to the interactive nature of our systemic observability techniques, we believe that there is room for automation to identify some of the more common performance anti-patterns. However, the packages - and there are dozens of them - are lacking in the ability to reliably and repeatedly exercise a significant portion of the application code base.

The products in this space have not kept pace with the increasing complexity of modern systems - clouds, grids, OS virtualization, and more. Surveying the market we have found most provide 2-dimensional data analysis - often focused on a single metric (Transactions Per Second for example) with wall clock time as the other axis. Some have very impressive looking charts and graphs but provide very little analysis of the data or indication of potential throttle points (e.g. multi-variable correlations, standard deviations, small interval outlier identification). Those that claim to provide these features in their tools require that you specify exactly what relationships you find interesting.

For the limited amount of useful data provided by these load profilers, there is a significant overhead in hardware resources - dozens of smaller servers to drive a large stage or production environment as well as your personnel's time. In addition to the high costs of acquisition, these on-going costs can become burdensome on already stringent budgets.

In addition to monitoring and analysis tools there is another important product space hindering the optimization of resources. The ability to execute repeatable test cases, whether for performance, scalability, regression or some other purpose, can be very challenging. Many organizations have made significant investments in software to assist in running tests suites and after some initial use they sit on the shelf as maintenance costs exceed the benefits realized. In most large organizations there is a small percentage that use these tools across their full application suite. For those application teams that do use these tools there is often a significant morale drain on the staff required to create and maintain these tests. Unfortunately, it is often a thankless job. A great deal of analysis is required to ensure the tests fairly represent the actual production workloads and create new tests as business and functionality grows.

At Aeysis we are currently developing software products to address each and every one of these requirements and more. While each of the tools can be used separately they will be tightly integrated to provide a very high value solution. With our product in the early design and implementation phases, we cannot provide specifics without appropriate non-disclosure agreements, but suffice to say: manually coding test cases and validating results while attempting to isolate the top potential causes for performance regression are a thing of the past.