PGConf.Russia 2023

The developer shares his experience of migration a highly loaded system in the field of railway transport from IBM DB2 for z/OS to PostgresPro in a multimaster configuration. The main focus is on the nuances of working with a cluster operating in multimaster mode, what problems had to be faced and how to solve them.

The importance of sharding is now colossal. The size of today's databases exceeds 100 terabytes, scaling up vertically, and adding replicas containing a complete physical copy of the database is becoming difficult, especially when computational resources are scarce. Database sharding is a way to scale horizontally by partitioning data between compute nodes that are independent of each other.

In the PostgreSQL world, there are well-known scaling tools - CitusDB, and Greenplum - as well as new generation solutions - Cockroach DB, Yugabyte DB, SPQR, and Shardman.

In this talk, we will discuss the difference between these implementations, the advantages, and disadvantages of these solutions, examine the current state of implementation of sharding in vanilla PostgreSQL, and also touch on another important topic - providing data integrity and consistency guarantees across a distributed cluster.

Every DBA has in one way or another experienced the situation when PostgreSQL tables and indexes grow significantly in size. While looking for the reason for such behavior, we often conclude that database objects have "bloated". In this talk, we'll discuss the reasons behind bloating, create a testing environment to define the best method to reduce bloating. We'll also compare several anti-bloating utilities and get familiar with one more tool that helps us to efficiently struggle against bloating. We expect this presentation to become helpful for PostgreSQL DBAs of any experience level.

I will talk about the common mistakes when setting up a Corosync-Pacemaker failover cluster. Often these mistakes lead to fatal consequences, and, as a result, to the rejection of the chosen solution in favor of others. Would you like to get a recipe for the "right" cluster?