I threw together a quick benchmark, and was seeing ~4-5x slower async calls vs sync (with the network mocked out)
This mostly appears to be due to re-building the wrapped function on each call, rather than caching. I opened googleapis/gapic-generator-python#1755 to address this
I threw together a quick benchmark, and was seeing ~4-5x slower async calls vs sync (with the network mocked out)
This mostly appears to be due to re-building the wrapped function on each call, rather than caching. I opened googleapis/gapic-generator-python#1755 to address this