We always care about how much workload our software can handle, especially in large-scale scenarios. However the reality is not always ideal, but almost always comes with skewed workloads and fluctuating resources, challenging our software’s stability and tail latency. In this session, I’ll share our experience on tuning a product-ready Rust system to achieve predictable scheduling and promising self-protection in complex environments. Including failed tries, what existing Rust infrastructure can’t achieve, and how we get there.

Speakers:

Ruihang Xia: Software Architect at Greptime Inc.

Software Architect at Greptime Apache DataFusion PMC, Apache Arrow Committer, Apache HoraeDB PPMC https://github.com/waynexia