Custom. Proprietary. Exclusive. All words we associate with quality and better performance custom suits fit better, proprietary software meets our specific needs, exclusive access grants us a social edge.
But is that always true? But is that always true? For vintage cars, probably. But, for an RDBMS? Not always. We put PostgreSQL 9.1 against a proprietary RDBMS and the results are in.
Using the Open Source load test tool, Hammerora, we simulated 500 users, each performing 10,000 transactions against 800 data warehouses and measured sustained throughput. The same platform was used for all tests and repeated multiple times (after reset) ensuring the repeatability of the results.
So, how did PostgresSQL do? In short, consistently better. During our testing, the average Proprietary RDBMS sample showed throughput of 53824 Transactions per Minute (TPM) while PostgreSQL samples averaged 64374 TPM, - an almost 20% higher throughput.
For the full results, (click here)
But is that always true? But is that always true? For vintage cars, probably. But, for an RDBMS? Not always. We put PostgreSQL 9.1 against a proprietary RDBMS and the results are in.
Using the Open Source load test tool, Hammerora, we simulated 500 users, each performing 10,000 transactions against 800 data warehouses and measured sustained throughput. The same platform was used for all tests and repeated multiple times (after reset) ensuring the repeatability of the results.
So, how did PostgresSQL do? In short, consistently better. During our testing, the average Proprietary RDBMS sample showed throughput of 53824 Transactions per Minute (TPM) while PostgreSQL samples averaged 64374 TPM, - an almost 20% higher throughput.
For the full results, (click here)