Skip to content

load-test.md

Latest commit

 

History

History
54 lines (40 loc) · 1.25 KB

load-test.md

File metadata and controls

54 lines (40 loc) · 1.25 KB

Revi Load Testing Guide

Overview

Revi no longer keeps a Python load-test stack. The supported performance checks now use:

  • cargo bench --manifest-path server/Cargo.toml for in-process Rust API benchmarks.
  • bash scripts/load-smoke.sh for concurrent real-server smoke traffic against the running Rust backend.

Prerequisites

# Start the supported Rust backend first
cd /path/to/Revi
cargo run --manifest-path server/Cargo.toml --bin revi -- --port 8000

Run Commands

API micro-benchmarks

cd server
cargo bench

Concurrent smoke load

bash scripts/load-smoke.sh

Optional tuning:

REVI_LOAD_REQUESTS=100 REVI_LOAD_PARALLELISM=8 bash scripts/load-smoke.sh http://localhost:8000

Targets

  • Smoke load should finish without HTTP failures.
  • cargo bench should stay stable across runs for GET /api/reviews, GET /api/reviews/{item_id}, POST /api/comments/{item_id}, and GET /api/export/{item_id}.

Notes

  • scripts/load-smoke.sh targets the real Rust server, not the in-memory test harness.
  • cargo bench measures handler-level performance without requiring any Python tooling.
  • For longer runs, monitor memory with top or htop while the Rust backend is serving requests.