Checkout the slides

Introduction.

Have you ever been in a database emergency and needed real quick pinpointing information on what is wrong with a PostgreSQL instance/database?
Are you looking for a good Analysis and meaningful conclusions using the data presented by PostgreSQL performance views (pg_stat_*)
Have you ever felt that there is too much performance and monitoring data and how to make sense of that?
If the answer is "Yes" for any or all of these. This is a presentation for you.
Content primarily targeted for DBAs and Support Engineers who need to deal with operational emergencies frequently. Understanding frequent issues will be helpful for designers and architects.

Contents

• Most Important information the PostgreSQL stats views give us is their indications.
• How to effectively utilize the wait-event information to understand the problems and performance bottlenecks
• How to analyze database-level information How to analyze table and Index information
• How to check autovacuum efficiency? What are the practical problems with autovacuum tuning?
• Make meaningful ideas about table and index information.
• Explaining major problems encountered by many users
• identification of indicators of poor tuning practice.
• What data we need to collect
• Data from Host OS and how to correlate.

This presentation also includes Sharing of Simple scripts and methods for data collection and analysis. I will be sharing some of the best practices, also.