What could possibly go wrong if you try to load 1 Petabyte into PostgreSQL
Presented by:
Michael Zhilin
Head of the performance group of the implementation and technical support department. He has been working on improving computing performance for more than 10 years. Now he is a DBMS performance expert in Postgres Professional and is professionally engaged in performance analysis, evaluation of optimisation works and configuration of PostgreSQL and Postgres Pro DBMS.
No video of the event yet, sorry!
At the end of year 2024, we decided to run a small, quick test of the PostgreSQL-based distributed database called Shardman on 1 petabyte of data. The plan was to rent the hardware, run one of popular benchmark, and prepare a report with the results. But from the very beginning, nothing went according to plan.
In this talk, we will share:
- The real story of a single benchmark test
- Our benchmarking methodology
- Issues with the hardware, the database, and the benchmark itself
- And the results...
If you have also run benchmarks or are just planning to do so, come along, and let's discuss the pain points, life hacks, and other philosophical questions of the non-quantum world together.
- Date:
- Duration:
- 45 min
- Room:
- Conference:
- PGConf India, 2026
- Language:
- Track:
- Case Study
- Difficulty:
- Medium