This talk looks at common and maybe not so common pitfalls in writing high-performance distributed systems in Rust. We will look at topics ranging from memory layout having an outsize impact on the overall latency in a distributed system, to reducing the memory footprint of scatter-gather operations, to writing custom SMID kernels to speed up compute-intensive parts of both read and write paths. Performance can be gained in surprising places, where often small changes have a big impact.