Talks

• Some theory: what AQO is and how it influences query planning
  
• Testing planning: selection criteria and test preparation
  
• Testing: what and how was measured
  
• Findings

A talk on how regular™ indexes work in PostgreSQL. The talk will be extremely useful for those who are starting to work with databases, and for those who have worked with them but have forgotten. If the talk hits the golden picks of "must-see onboarding developer", then the goal is achieved. We will tackle the following subtopics:
- How an index can speed up search;
- Should we index WHERE conditions;
- Should we index ORDER BY statements;
- Should we index foreign keys;
- What to do if search pre-assumes multiple conditions;
- Which column order should be specified in an index;
- Cases when index causes a slow-down and how to reduce its impact on the app.

PostgreSQL 16 is expected to be released in September. We are looking forward to a number of interesting and important changes in monitoring, logical replication and access management. My talk will cover these new features and many more; I am also open to conversations after the presentation.

Unfortunately, ideal computer systems exist only in science fiction books. PostgreSQL is not exception and sometimes problems may occur. I would like to discuss how to correctly (and incorrectly) try to solve a problem, which way to choose, which tool to use.

The talk is of interest to both beginners and experienced users and database administrators.

We started using PLV8 in development in 2021. Primarily, we have started to build our Postgres + our own needs. However, after some time, we found an active growth in downloads. Over the past 10 months, our image has been downloaded over 100 thousand times, while we have not been engaged in its promotion. In this report, I will try to give an understanding of the growth in popularity of the project, briefly talk about its development, which improvements have aroused interest in the community, which have not, and share plans for the future.

Traditionally, fault tolerance in Postgres is achieved using built-in replication mechanisms and external utilities that monitor the state of running Postgres instances and react accordingly when various failures occur. In my presentation, I will tell you what we like and what we don’t like about this approach, what alternative we see, what we have been able to achieve to date and what we want to get by the release, which is planned for December.