PgConf.Russia 2016

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.

Common Java wisdom is to use PreparedStatements and Batch DML in order to achieve top performance. It turns out one cannot just blindly follow the best practices. In order to get high throughput, you need to understand the specifics of the database in question, and the content of the data.

In the talk we will see how proper usage of PostgreSQL protocol enables high performance operation while fetching and storing the data. We will see how trivial application and/or JDBC driver code changes can result in dramatic performance improvements. We will examine how server-side prepared statements should be activated, and discuss pitfalls of using server-prepared statements.

When new versions of Postgres are released most of the attention is focused on new features. Inevitably a release note claiming speed improvements seems relatively mundane and doesn't provide the compelling argument for upgrading. However the reality is that these speed improvements represent pain points that have been identified and solved.

Reviewing the changes to the sort code in Postgres over the last 10 years clearly shows the kinds of problems users have run into. As usage patterns changed over years, databases scaled up, and hardware changed new problems arose and drove further development to solve them.

Upcoming changes in 9.5 and 9.6 will dramatically change the experience further. Making sorting UTF8 and other encodings less of a problem and handling scaling to larger machines with many processors and memory cache more effectively.

Postgresql's buffer manager has parts where it's showing its age. We'll discuss how it currently works, what problems there are, and what attempts are in progress to rectify its weaknesses.

• Lookups in the buffer cache are expensive
• The buffer mapping table is organized as a hash table, which makes efficient implementations of prefetching, write coalescing, dropping of cache contents hard
• Relation extension scales badly
• Cache replacement is inefficient
• Cache replacement replaces the wrong buffers