Linux load average and other silly metrics

Linux load average and other silly metrics

Today’s prevalent operating system is Linux, where Oracle Database runs with buffered or direct I/O. Performance monitoring relies on metrics such as %CPU, load average, I/O wait, and free memory.
While these metrics are easy to display, interpreting them correctly can be challenging.

In this presentation, we will explore the intricacies of performance monitoring by running a simple workload and examining the “top” command together.
We will uncover the limitations and misinterpretations when relying solely on these metrics.

About Franck

Franck is a Developer Advocate for YugabyteDB, an open-source, PostgreSQL-compatible, distributed SQL database.

With over 20 years of consulting experience, he brings deep expertise and a genuine passion for databases. Franck holds certifications such as Oracle Certified Master, AWS Database Specialist, and has been recognized as an Oracle ACE Director, AWS Data Hero, and AWS Academy educator.

Beyond his role, Franck actively shares his knowledge and engages with technical communities.

You can connect with him on Twitter and LinkedIn for insightful discussions and updates. His writings on various database topics can be found at https://dev.to/franckpachot, where he shares valuable insights and perspectives.