Developing a distributed queue with delayed tasks based on PostgreSQL

Maksim will show the process of creating a task queue based on a relational database. It supports retries and executing tasks in strict sequence by one entity. The slides will show code in Java and SQL: the structure of tables and rows. The method itself is based on Postgres locks and waiting for locks within a transaction while the task is running. The speaker will also talk about statistics, how much the queue weighs, its indexes and where this data comes from (a bit of Postgres troubleshooting training). And at the end, some load results as well.