Skip to content

TradAI Troubleshooting Guide

Quick diagnosis and resolution for common issues in the TradAI platform.

Quick Diagnosis

Find your symptom and jump to the solution:

Symptom Likely Cause Guide
"Connection refused" Service not running Common Issues
"Authentication failed" Token/Cognito issues Common Issues
"Backtest stuck" ECS task failed Debug Workflows
"Data not fresh" Sync job failed Service Interactions
"Strategy not found" Config missing Common Issues
"MLflow error" Server unavailable Service Interactions
"Timeout errors" Network/NAT issues Common Issues
"Out of memory" Insufficient resources Debug Workflows

Guides

Guide Description
Common Issues FAQ-style solutions for frequent problems
Debug Workflows Step-by-step debugging procedures
Service Interactions Inter-service communication issues

Quick Commands

Check Service Health

# All services
just doctor services

# Individual service
curl http://localhost:8000/api/v1/health | jq

# ECS services (AWS)
aws ecs describe-services \
  --cluster tradai-${ENVIRONMENT} \
  --services tradai-backend-${ENVIRONMENT} \
  --query 'services[0].{Status:status,Running:runningCount,Desired:desiredCount}'

View Logs

# Local (Docker)
docker compose logs -f backend
docker compose logs -f strategy-service

# AWS CloudWatch
aws logs tail /ecs/tradai/${ENVIRONMENT} --follow  # All services share one log group

Check Running Tasks

# Docker
docker compose ps

# AWS ECS
aws ecs list-tasks --cluster tradai-${ENVIRONMENT}

Test Connectivity

# Local services
curl http://localhost:8000/api/v1/health
curl http://localhost:8003/api/v1/health
curl http://localhost:8002/api/v1/health

# AWS (via API Gateway)
curl https://${API_ENDPOINT}/health

When to Escalate

Scenario Action
Single failed request Check logs, retry
Repeated failures (>3) Check service health, investigate
All services affected Check infrastructure (VPC, NAT, DNS)
Data inconsistency Check DynamoDB, S3, pause operations
Security concern Follow Security Incidents