Substantially slower throughput recorded while running native (16k I/O-size) Progress OpenEdge workloads versus synthetic benchmarking results done with same (16k) I/O-size on vSAN.
VMware vSAN
The performance difference is a result of synchronous processing of I/O's used by Progress DB engine.
NOTE: This is default (and only supported) mode, according to Progress DB documentation:
Does Progress use Synchronous or Asynchronous I/O?
This synchronous requirement creates a substantial bottleneck, as it prevents the application from utilizing the high-concurrency capabilities of vSAN.
Before migrating any Progress workload to a vSAN enabled cluster, it is advisable to benchmark the environment accordingly to confirm the maximums.
Please engage with VMware PSO (Professional services) to evaluate how to best optimize this specific database configuration within the existing infrastructure and greenfield deployments.