Performance Testing

EMC's performance testing utilizes and allows for agile software delivery methodologies. We strive to eliminate any testing bottlenecks for our clients and reduce to minimum their overhead supporting them in the challenge of increased cost of implementation, integration and maintenance of the various IT solutions in their organizations using different standards, formats and frameworks. Our testing services are integrated within the development lifecycle providing for optimal software quality, product delivery, risk and cost efficiency as well as time to market. In line with all EMC service offerings, the work can be conducted on-site with our Professional Services consultants, or off-site at EMC’s Test Lab.

Performance testing from EMC includes four core non-functional tests:

Stress test

Stress test aims to determine under what condition an application will fail, how it will fail and what indicators can be monitored to warn of an impending failure.

Value Added:

Evaluate and determine the maximum load limit of an application before performance issues, failures and errors occur
Find out if any data corruption can be achieved via overstressing the IT solution
Guarantee that overstressing the IT solution will not provide for any security loopholes and failures
Identify the implications of and dependencies to unavailability of supporting hardware and applications
Determine and categorize most important risks, critical issues and failures to prevent and handle during peak loads for the relevant solution
Define and create pro-active application and system monitoring in order to notify in advance about possible issues and avoid them

Tools:

HP Mercury LoadRunner; IBM Rational Performance tester; IBM Rational Robot; Borland SilkPerformer; RadView Webload; Compuware QALoad; Oracle Application Testing Suite; Oracle Solaris Studio; JMeter; OpenSTA ; Grinder.

Soak (Endurance) test

Soak (Endurance) test aims at identifying and confirming performance patterns of the solution under test when exposed to workload models and load volumes expected during live operations over a prolonged timeframe.

Value Added:

Test available storage and hardware capacity and detect limits before possible errors and failures occur
Identify memory issues
Discover any performance degradation factors (increased amount of data, amount of application processes, CPU and RAM usage, storage, DB connections, dependencies on other processes on the server, etc.)
Determine consistency of the performance and availability of the tested hardware and software over time
Identify any problems accumulating and emerging over time

Tools:

Capacity testing

Capacity testing defines the boundaries of your hardware and the final limit at which it will fail. It is performed in combination with capacity planning identifying possible scalability and optimization needs taking into account current and future usage, users, data volumes, performance and time requirements. For example in order to enable the new projected utilization of a respective IT solution one might need to add and plan for various additional hardware as well as upgrading the characteristics of available one (new application or Citrix servers, more CPUs, Database and hard drive space, RAM, virtual memory, network throughput, etc.)

Value Added:

Define present capacity of the system in test and evaluate its utilization
Test capacity scenarios and provide actual data for optimal planning
Validate or reject various future capacity setups and models

Tools:

Spike test

Spike test aims at identifying and confirming performance patterns of the solution under test when exposed to workload models and load volumes increased beyond those of live operations over multiple short timeframes.

Value Added:

Discover any performance degradation factors and the ability of the system to stabilize after a short peak load
Identify memory issues
Determine consistency of the performance and availability of the tested hardware and software in extreme load situations
Describe and categorize the identified issues in such situations
Define indicators and possible pro-active monitoring in order to avoid and prevent such issues

Tools:

Our Methodology

Through EMC's proven methodology, we provide our clients with lower cost for their performance testing projects as well as optimal value and quality for their investments. Our services help you identify early enough any bottlenecks in your respective IT solutions and enterprise architecture thus allowing for flexibility and avoiding major expensive and complicated changes related to them in the future.

EMC's professionals use generic agile methodology for Performance testing providing the ability to integrate it into various stages of the software development process. The following diagram describes the key steps involved:

web by SP Vision Ltd.