Performance concepts and configuration

WWR2 : Performance concepts and configuration

You create workloads in Storage Modeller to assess the performance of your models. Workloads can ingest real-world data that you import into Storage Modeller. Such data can maximize the confidence of your performance assessments.

The topics in this section describe the following aspects of performance assessment in Storage Modeller:

▪“Introduction to performance modeling in Storage Modeller”

– “Some ways that configurations affect performance predictions”

– “Interpreting the response-time curve”

– “Computation of performance predictions”

– “Performance terminology”

▪“Importing data for performance assessments”

– “Process overview for data ingestion”

– “Imported Data perspective”

– “Importing a data record”

– “Getting IBM i data to import”

– “Getting IBM Z data to import”

▪“Ingesting data into a workload”

– “Orientation points”

– “Process review: import and ingest”

– “Ingestion procedure”

▪“Workloads for performance assessment”

– “Data collection”

– “Getting the best possible statistics for workloads”

– “Validations in the Workload form”

– “Continuous improvement for Workloads: Unsolved, Solved, and Stale”

▪“Adding a workload”

– “About Advanced Options in the Workload form”

– “How would a bigger cache affect my new system?”

A key sizing consideration and a Note

For IBM Storage Virtualize systems, Storage Modeller optimizes your system configurations to fully utilize all available resources, such as CPU cores. When you assess performance of your design, your workloads are spread evenly across all ports, CPU cores, and so on. As a result, you get optimal performance reports for the system.

Note: Some customers present design requirements that are counter-productive. For example, they assume that a small number of very large volumes is viable. IBM users must use Storage Modeller with an awareness of this type of issue and apply best practices for sizing projects whenever possible. Then, customers can more easily be persuaded to optimize their requirements.