PGConf.Russia 2022

Previously, we have explained the internals of Avito, discussed, where and how we store your classifieds, and how they appear on the search results. In the recent 3-4 a lot has changed in Avito. We got rid of logical replication, stopped using standby servers for reads, removed nearly all stored procedures and our custom failover solution, migrated all our classifieds into a sharded DBMS, switched from the monolith app to microservices. I'll explain why we made such decisions, list some of the problems we encountered and describe the current state of our development process.

Replication is one of the important mechanisms designed to provide database fault tolerance and scalability. In our practice, we regularly encounter the problem of low replication performance. This prompted us to investigate the factors that affect the speed of physical replication. In this presentation, I will talk about our findings. I will also demonstrate the differences in replication performance for various versions of PostgreSQL.

During my career, I've seen a lot of code, and very often I faced inaccuracies in handling dates and times. Sometimes the parties got discrepancies in the monthly reports due to such inaccuracies; or daily reports were different for Moscow and SFO, etc. I wouldn't call it a serious problem, but it's annoying and time-consuming. In all cases, such issues occur because of neglectful handling of dates and times. In my presentation, I will discuss how we can avoid it.

My talk covers PostgreSQL infrastructure in Ozon: - how do we tune virtual machines (KVM); - which version control system we have created; - which drivers we make and how it relates to fault tolerance and load balancing; - how do we make "hot" upgrades for the parameters of our virtual machines.