gavindiaz/polyfill-rs
1
0
Fork 0
mirror of https://github.com/floor-licker/polyfill-rs.git synced 2026-08-24 01:48:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: integrate infrastructure modules and achieve 8.9% performance improvement

Browse Source 
Integrated DNS caching, connection manager, and buffer pool modules into ClobClient structure to enable production use. Added start_keepalive() and stop_keepalive() methods for maintaining warm connections through background keep-alive pings. Implemented DNS cache pre-warming on client initialization and buffer pool with 512KB buffers for reducing allocation overhead. Cleaned up temporary test and analysis files from optimization exploration. Benchmark results with keep-alive enabled show 368.6ms mean latency compared to polymarket-rs-client's 404.5ms, representing 8.9% improvement and 35.9ms faster performance while maintaining production-safe approaches
This commit is contained in:
floor-licker
2025-12-06 17:47:35 -05:00
parent af9e1a1939
commit dc6092c244
13 changed files with 206 additions and 1392 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+6 -4
View File
@@ -50,12 +50,14 @@ End-to-end performance with Polymarket's API, including network latency, JSON pa
| Operation | polyfill-rs | polymarket-rs-client | Official Python Client |
|-----------|-------------|----------------------|------------------------|
| **Fetch Markets** | **382.6 ms ± 75.1 ms** | 404.5 ms ± 22.9 ms | 1.366 s ± 0.048 s |
| **Fetch Markets** | **368.6 ms ± 67.1 ms** | 404.5 ms ± 22.9 ms | 1.366 s ± 0.048 s |
**Performance vs Competition:**
- **5.4% faster** than polymarket-rs-client (Rust) - 21.9ms improvement
- **3.6x faster** than Official Python Client
- **8.9% faster** than polymarket-rs-client - 35.9ms improvement
- **3.7x faster** than Official Python Client
**Note:** Best performance achieved with connection keep-alive enabled (`client.start_keepalive(Duration::from_secs(30)).await`).
**Computational Performance (pure CPU, no I/O)**
@@ -67,7 +69,7 @@ End-to-end performance with Polymarket's API, including network latency, JSON pa
**Key Performance Optimizations:**
polyfill-rs achieves 5.4% better performance than polymarket-rs-client through several targeted optimizations. We use simd-json for SIMD-accelerated JSON parsing, which provides a 1.77x speedup over standard serde_json deserialization and saves approximately 1-2ms per request. Our HTTP/2 configuration has been empirically tuned through systematic benchmarking, with a 512KB initial stream window size proving optimal for the typical 469KB payload sizes from Polymarket's API. We've implemented DNS caching to eliminate redundant lookups, connection keep-alive management to maintain warm connections, and a buffer pool to reduce memory allocation overhead during request processing. These optimizations collectively reduce mean latency from 401ms to 382.6ms while maintaining production-safe, conservative approaches.
polyfill-rs achieves 8.9% better performance than polymarket-rs-client through several targeted optimizations and infrastructure integration. We use simd-json for SIMD-accelerated JSON parsing, which provides a 1.77x speedup over standard serde_json deserialization and saves approximately 1-2ms per request. Our HTTP/2 configuration has been empirically tuned through systematic benchmarking, with a 512KB initial stream window size proving optimal for the typical 469KB payload sizes from Polymarket's API. The client includes integrated DNS caching to eliminate redundant lookups, a connection manager with background keep-alive to maintain warm connections (preventing costly reconnections), and a buffer pool to reduce memory allocation overhead during request processing. These optimizations collectively reduce mean latency from 401ms to 368.6ms (with keep-alive enabled) while maintaining production-safe, conservative approaches.
**Performance Breakdown:**
- Network (DNS/TCP/TLS): ~150ms (optimized with DNS caching and HTTP/2 tuning)