Question
Hi @author, thank you for the great work on sideway and the detailed blog post!
I read the performance comparison between stride and perftest and noticed that
bandwidth/latency results are very close. I'm interested in understanding the
CPU and memory resource usage during these benchmarks.
Specific questions:
- Did you measure CPU utilization (e.g., via
perf, top, or pidstat) during
the bandwidth/latency tests? If so, how do stride and perftest compare?
- What about memory footprint? Does the Rust wrapper introduce any noticeable
overhead in heap allocations or RSS usage?