Capacity planning
While, on the one hand, a computer system must meet very specific performance requirements for its users, on the other hand, the company must control the costs of the IT infrastructure.
For this reason, periodic activity must be carried out Capacity Planning that ensures correct system sizing as application operating scenarios change. The aim of the course is to provide a effective methodology and to show the most suitable tools for carrying out Capacity Planning periodically.
The CP process begins with a systematic measurement of performance and load volumes, moves through the parameterisation of an easily “intuitable” mathematical model, and concludes with an equally systematic activity of What-if analysis, which mainly allows you to:
- assess the maximum system capacity; For example, to estimate the maximum number of users the system can handle
- compare the system's maximum capacity with the trend curves of the load to predict if and until when the system will be able to operate within the set operating limits, evaluate the impact on performance following fail-over
- identify the resources and systems which constitute a bottleneck for performance so as to plan any upgrades
- assess the performance impacts in buffer applications, such as, for example, the size of the connection pool to a database server.
Contents
Performance of an IT system, response time and application throughput; what is capacity planning: system model and workload model.
Resources and requests, resource classification (queued, delayed, multiservice, passive), load model and visits, parameter calculation.
Open and closed systems, approximate techniques and exact techniques, bottleneck identification.
Definition of objectives, analysis of the architecture and application, construction of the system model and load model, activation of collection agents, calculation of model parameters, resolution of the model.
Maximum and residual capacity of a system, bottleneck removal, analysis of a system in case of failure or upgrade.
Prerequisites
Participation in the course “Measuring and maintaining Web application performance”. General knowledge of systems management issues.
The course does not require any particular mathematical background.