Volume/Performance/Capacity Testing
This is designed to ensure that the hardware and networking environments will support the software and a typical processing workload with a normal level of concurrent users using various system functions and background processes and that major business processes, transactions, enquiries and reports perform in a reasonable timeframe without extravagant impact on machine resource usage.
Standard forms of benchmarking can provide detailed technical information on the hardware and software environments being used for testing, but the information is often inadequate as a guide as to how the system will really perform with a full set of real users. Therefore, rigorous Volume and Performance Testing are vital activities prior to going live.
For the testing to be entirely credible, the new system would have to operate in the system configuration that will be used for operational purposes post go-live and be tested by a typical number of users accessing the system at a popular time from their natural locations and with an average make-up of machine activity and other system processing also taking place. However, this is not always feasible (for example the requisite number of people may not have been trained at the point of undertaking these tests), in which case alternative arrangements will have to be made to make best use of the system configuration that is available, simulated machine activity and system processing and a mixture of actual and automated user access.
During System/Functionality Testing, when no provisions have been made specifically to measure volume or performance capabilities, users should record response times for their normal transactions and processes and bring to the attention of the functional and technical consultants those, if any, showing significantly poor performance compared with their experiences of the old system(s).
The same users should then compare times of running the same transactions on the new system when Volume and Performance Testing provisions are in place, and record their findings, again for review by the functional and technical consultants.
Where performance remains poor, the functional and technical consultants should review the specific poorly performing processes, transactions, enquiries and reports to see whether improvements can be made via the configuration. Repeat performance testing should take place so that users can check whether the requisite improvements have been made.
Each user involved in Volume and Performance Testing and relevant technically qualified personnel should be asked to report their findings in accordance with the institution's project procedures and confirm whether the major business processes, transactions, enquiries and reports perform in a reasonable timeframe without extravagant impact on machine resource usage. Also, measurements of the computer's overall performance should be taken and documented and reported in the same way. Technically qualified personnel should then be asked to review those measurements and confirm whether the capacity of the planned system configuration is indeed sufficient to support the software and a typical processing workload, or would be sufficient to support them when the planned environments and typical workload are in place.
