PGConf.Russia 2018

Rambler & Co is a lot of publications, services and projects. Appear new and grow existing. This environment requires a reliable, fault-tolerant, scalable, automated system.

I'll tell you about the structure of our PostgreSQL SaaS, what tools and technologies we use. Quorum of 3 Data Centers. A single entry point for clients based on dynamic routing. Emergency switching of the primary server. Transparent scaling for reading. Create a replica without load on the cluster. Transparent transfer of PostgreSQL cluster to other servers. Update dev environment from prod for development. Backup with compression and the use of multiple CPUs on the side of the database, the restoration of one database from basebackup. Monitoring of sql queries.

DBeaver is the Universal Databases Manager. We creating DBeaver using Open Source Software model. We pay special attention to the PotgreSQL because of its capabilities, popularity and OSS nature. We will talk about DBeaver evolution and structure, will demonstrate the basic functionality. We are going to cover challenges of PostgreSQL client creation. Also we will discuss requirements management model and interaction with the PostgreSQL community. We will show in details how to debug the PL/pgSQL stored procedures interactively. We are going to introduce some new features of the nearest DBeaver release and to share our plans.

«KNOPKA» («The button»), the largest outsourcing of accounting, will tell you about your experience of transferring 500Gb of 1C Fresh databases to PostgreSQL. The report touches on the choice of DBMS and the fight against stereotypes, where we started and how developed the use of PostgreSQL, the approaches to backup and recovery of information security, our current performance of DBMS and 1C, a look into the future. Let us share why we believe that 1C + PostgreSQL is simple, reliable and fast.

PostgreSQL looks very competitive with other mainstream databases on OLTP workload (execution of large number of simple queries). But on OLAP queries, requiring processing of larger volumes of data, DBMS-es oriented on analytic queries processing can provide an order of magnitude better speed. The following factors limit Postgres OLAP performance:

• Unpacking tuple overhead (tuple_deform)
• Interpretation overhead (Postgres executor has to interpret query execution plan)
• Abstraction penalty (support of abstract data types)
• Pull model overhead (operators are pulling tuples from heap page one-by-one, resulting numerous repeated accesses to the page)
• MVCC overhead (extra per-tuple storage + visibility check cost)

All this issues can be solved using vectorized executor, which proceed bulk of values at once. In this presentation I will show how vector operations can be implemented in Postgres as standard Postgres extension, not affecting Postgres core. The approach is based on introducing special types: tile types, which can be used instead of normal (scalar) types and implement vector operations. Postgres extension mechanism, such as UDT (user-defined type), FDW (foreign data wrappers), executor hooks are used to let users work with vectorized tables almost in the same way as with normal tables. But more than 10 times faster because of vector operations.