SOROBAN_BATCH_READS.md

Soroban Batch Contract Reads Optimization

Overview

This document describes the batch contract read optimization implemented in SorobanQuestReaderService.

Problem Statement

Sequential Soroban RPC calls for multiple on-chain entities (e.g. fetching status for 20 quests) accumulate network round-trip latency linearly ($O(N)$), creating heavy user-facing response delays.

Solution Architecture

SorobanQuestReaderService.getQuestsBatch(contractId, questIds, options) executes contract simulation reads concurrently with bounded concurrency chunking.

Environment Configuration

• SOROBAN_BATCH_READ_CONCURRENCY: Default concurrency limit per batch (Default: 10).

Benchmark Results

Running npx ts-node scripts/benchmark-soroban-batch-reads.ts:

• Sequential Reads (20 quests): 423ms total duration (21.15ms per quest).
• Batched Reads (20 quests, concurrency=10): 43ms total duration (2.15ms per quest).
• Impact: 89.8% latency reduction, 9.84x speedup factor.