PgConf.Russia 2016

Schema-less is definitely a trend in the data storage nowadays, and it's not only about NoSQL, but also about traditional RDBMS. Many relational databases (e.g. PostgreSQL, Oracle, db2, Mysql) allow to storing data in the schema-less json format and use their own more or less unique way to do that.

This talk contains two parts:

• Comparison of the json support in PostgreSQL and different relational databases, namely Mysql, Oracle, db2, MSSql in terms of supported features, functions and so on.
• Performance benchmarks for databases with the advanced json support, namely PostgreSQL and Mysql, and the MongoDB on different workload types and configurations.

Multicorn is a generic Foreign Data Wrapper which goal is to simplify development of FDWs by writing them in Python.

We will see:

• what is an FDW what Multicorn is trying to solve how to use it, with a brief tour of the FDWs shipping with Multicorn.
• how to write your own FDW in python, including the new 9.5 IMPORT FOREIGN SCHEMA api.
• the internals: what Multicorn is doing for you behind the scenes, and what it doesn't

After a presentation of FDWs in general, and what the Multicorn extension really is, we will take a look at some of the FDWs bundled with Multicorn.

Then, a complete tour of the Multicorn API will teach you how to write a FDW in python, including the following features:

• using the table definition
• WHERE clauses push-down
• output columns restrictions
• influencing the planner
• writing to a foreign table
• IMPORT FOREIGN SCHEMA
• ORDER BY clauses pushdown
• transaction management

This will be a hands-on explanation, with code snippets allowing you to build your own FDW in python from scratch.

We'll discuss how does Linux work with virtual memory. The following topics will be covered: * x86-64 page table, context switch and page fault; * internals of virtual memory management (VMM) in Linux; * page eviction methods in Linux, page cache and anonymous pages; * huge and gigantic pages, transparent huge pages; * how mmap(2) works and what madvise(2), msync(2) etc. provide; * why large databases don't use mmap(2), but rather implement buffer pool on their own; * ans surely how to tune Linux VMM using sysctl.

We love to stress test software, since we are a performance engineering company. Our friends from a hosting company servers.com provided us with a modern dedicated server so we immediately started to test PostgreSQL in different environments, including SmartOS, DragonFly and Windows. We would like to present our results (and all the gory details) to community.