Acumatica DB Benchmark Part 2: PostgreSQL Wins on Joins | MS SQL vs PostgreSQL vs MySQL

An update to our Acumatica database benchmark series. This time with a closer look at the test code and more detailed results. ▶ Part 1 — First benchmark results & visualization: https://youtu.be/lOD7-jrR86A?si=CFHs7yQeMJOWNOUW Yuriy Zaletskyy walks through exactly what each benchmark test measures, why the queries are non-trivial, and where the results get interesting. Who this is for: Acumatica developers, solution architects, and IT decision-makers who need real data to choose or justify a database engine for their ERP environment. 📋 What's tested (and why it's non-trivial): → Custom table read & write queries → Update & delete: non-standard implementations → Complex BQL join: Inventory Item + Item Class + Inside Status + GL Branch → PXProjection: custom projection built to test multi-table flat-view performance → Checksums used to verify data integrity across all runs → 5,000 & 15,000 records · 5 iterations per test 📊 Winner breakdown: → Microsoft SQL Server wins on: sequential read, write, update, delete · parallel write, update, delete → PostgreSQL wins on: complex BQL joins · PXProjection sequential · parallel read → MySQL 8.0 is the outsider across most categories The surprise: PostgreSQL outperforms SQL Server on complex joins and parallel reads, which matters if your Acumatica workload is join-heavy or runs concurrent queries. 🛠️ All code is open source. Run it on your own instances and compare. 📁 GitHub: https://github.com/yuriycto/AcumaticaDatabasesPerformanceAnalysis ───────────────────────── 🌐 AcuPower — Certified Acumatica Gold Partner ERP Implementation · Custom Development · Expert Delivery · Cloud ERP Migration #acumatica #acumaticaerp #postgresql #sqlserver #mysql #erpperformance #DatabaseBenchmark #BQL #PXProjection #acupower #Acumatica26r1 #opensource