Allocating Less: Really Thin Rust Cloud Apps
summary
Argues that lowering allocations is key for tiny cloud units and gives concrete tactics: measure with Dhat to find hotspots (example: Serde JSON value deserialization shows ~90% of allocations), replace per-entry heap strings to cut the ~72‑byte/hashmap entry overhead with crates like nested or small-string types, use jemalloc or mimalloc for faster alloc patterns, turn trait objects into enums via enum_dispatch, and prefer zero-copy formats (FlatBuffers/CapnProto/Arrow/scroll).
AI-written summary · about
chapters (6)
- 0:00Introduction and welcome
- 3:01Benefits of Rust for cloud
- 7:31Memory basics and profiling
- 12:43Strategies to reduce allocation
- 24:18Serialization and advanced tools
- 32:54Q&A session
related talks
26mAllocating Less: Really Thin Rust Cloud Apps
43mLow-level optimization of algebraic and similar structures
29mWrangling memory leaks where you thought they could never happen
31mReally Small Rust
01h 03mTaming dynamic memory
25mType System Tips for the Real World
32mBuilding and Maintaining Rust at Scale
19mRust Without Tears: A Guide to Avoiding Common Traps