This section shares the results of performance testing of the DBaaS platform in the Dell Technologies lab. We used HammerDB for these benchmarking efforts.
Our testing methodology considered all the architectural layers of the DBaaS platform holistically. Rather than stressing an individual SQL MI to its resource limits, we wanted to find a balance between SQL MI deployment density and cluster processor and memory utilization. We sought to prove the primary benefits of the DBaaS platform - rapid deployment, ease of management, and elastic resource scalability - while also optimizing cluster resource utilization and delivering performance that meets SLAs. To generate the appropriate load on the SQL MIs, we followed the industry-accepted practice of running an Online Transaction Processing (OLTP) workload from HammerDB.
We created a robust T-SQL logging and reporting framework to harvest HammerDB transactions per minute (TPM) metrics from a jump host outside of the DBaaS environment. The HammerDB CLI, with the associated .TCL scripting, was constructed to manage the autopilot-timed execution. This allowed us to properly capture TPM metrics from a multi-instance SQL workload at SQL MI instance granularity with individualized batch request/second logging.
With an Azure Arc-enabled SQL Server Managed Instance, there is no SQL host to deploy HammerDB against. However, SQL MIs run a complete SQL engine deployment, so we used the system databases to log metrics to a single table. Then we pulled these metrics back to a master reporting SQL instance for aggregation and review.
The metrics that are documented in the following testing section are aggregated TPM numbers for each SQL MI and user count for the entire run’s duration.
The following figure shows our testing topology architecture.
Figure 20. SQL MI testing topology