vichry2 commented on issue #6670: URL: https://github.com/apache/arrow-rs/issues/6670#issuecomment-2453475738
Hi @alamb, Thanks for the flamegraph tutorial! I've generated a flamegraph for my Rust server which I've attached here:  You can also find it on the `flight-benchmark` repo (`assets` folder). As mentioned previously, it appears that the bottleneck is related to memory movement (`memmove`) during batch encoding. Additionally, I’ve added commands in the `README.md` file of my repo that explain how to run the `client.py` script on the servers. This script compares the average response times for serving a single request in both languages, showing that Python is quicker. Regarding the screenshot of the heavy CPU usage during the Locust load test, I can get that to you within the week. I prefer using `taskset` to separate client CPUs from server CPUs, so you can see the server CPUs saturating before the client CPUs using `htop`. However, on my home computer (MacOS), I haven't found an easy equivalent for that. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
