closes #429: Add timeout and circuit breaker for external blockchain RPC calls

[sherifatolanike]

Description

Implements configurable timeout and circuit breaker pattern for external blockchain RPC calls to prevent cascading failures.

Changes

New files: backend/services/rpc/

• rpcConfig.ts — Configuration types & defaults for RPC endpoints with timeout, circuit breaker settings, and provider fallback ordering per chain
• circuitBreaker.ts — Full circuit breaker state machine (CLOSED → OPEN → HALF_OPEN → CLOSED) with events and typed errors
• rpcProviderFallback.ts — RPC provider with automatic fallback, AbortController-based timeout, and graceful degradation (cache_stale / fail_fast)
• rpcMonitorService.ts — Central monitoring dashboard, Prometheus metrics, per-chain health summaries, recent event log
• index.ts — Module exports

Modified files

• backend/services/shared/apiResponse.ts — Added 4 new RPC error codes
• backend/services/index.ts — Exported all RPC module types and classes
• backend/server.ts — Added 5 dashboard endpoints + provider initialization at startup

New API Endpoints

Endpoint	Description
GET /rpc/dashboard	Full circuit state dashboard with chain/endpoint filtering
GET /rpc/health/:chainId	Per-chain health summary
POST /rpc/reset	Manual circuit reset/open/reset_all/reset_metrics
GET /rpc/events	Recent circuit state change events
GET /metrics/rpc	Prometheus-formatted circuit metrics

Error Codes

• RPC_TIMEOUT (504) — Request exceeded timeout
• RPC_CIRCUIT_OPEN (503) — Circuit breaker is open, request blocked
• RPC_ALL_PROVIDERS_FAILED (503) — All RPC providers exhausted
• RPC_PROVIDER_NOT_FOUND (404) — No provider configured for chain

Technical Details

• Configurable timeout per RPC endpoint (default: 10s)
• Circuit breaker: 5 consecutive failures → OPEN, 30s recovery timeout → HALF_OPEN, 3 max probe requests
• Provider fallback with ordered endpoint list per chain
• Graceful degradation via stale cache responses when all providers fail
• Manual circuit reset via API for operations intervention
• Prometheus metrics integrated into existing monitoring infrastructure

Closes #429

[drips-wave]

@sherifatolanike Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits