Chapter 2. Best Practices Benchmarking RHEL-OSP 7 Environments

With a successful deployment completed as described in Deploying Red Hat Enterprise Linux OpenStack Platform 7 with RHEL-OSP director 7.1, the next steps involve benchmarking the RHEL-OSP environment. Step one involves using Rally as a benchmarking tool to capture initial baseline results. Baseline results are critical because they provide a foundation on how the RHEL-OSP 7 environment is performing. Without it, modifications that are made to the RHEL-OSP environment cannot be compared in order validate whether future tuning changes effected an environment positively or negatively. Once the baseline results for Chapter 3, Rally have been captured, step two involves the use of Chapter 5, Tuning the RHEL-OSP 7 Environment using Browbeat. Browbeat was originally developed to investigate the number of OpenStack Service workers needed and database connections utilized. It has since developed new features which allows us to check RHEL-OSP Deployments for common mistakes and run a variety of Rally scenarios in a automated fashion.

This paper focuses on using Browbeat specifically for its ansible playbooks that perform performance checks across the overcloud environment. This reference environment uses the performance checks to determine what is to be tuned within the RHEL-OSP 7 environment.

The following is a breakdown list of how to address benchmarking and tuning of a RHEL-OSP environment.

  • Benchmark using Chapter 3, Rally to capture baseline results.
  • Run performance checks on the RHEL-OSP environment using Browbeat
  • Re-run the Rally scenarios to capture latest results with the tuning changes.

The subsequent chapters provide the details in running the Rally benchmark and performance checking with Browbeat.