railway-logs（铁路日志）

Railway Logs

Access and analyze Railway logs for debugging and monitoring applications.

Overview

Railway provides comprehensive logging across the deployment lifecycle:

• Build Logs: Build process output and compilation
• Deploy Logs: Deployment lifecycle and health checks
• Runtime Logs: Application stdout/stderr output
• HTTP Logs: Request metadata (status, path, IP, timing)

All logs support filtering, search, and can be exported to external systems for long-term retention.

Keywords: Railway logs, build logs, deploy logs, runtime logs, HTTP logs, log filtering, log search, debugging, monitoring, observability

When to Use This Skill

• Debugging deployment failures
• Investigating runtime errors
• Monitoring application behavior
• Analyzing HTTP request patterns
• Filtering logs for specific events
• Exporting logs to external systems (Datadog, Axiom, BetterStack)
• Understanding build/deploy process
• Troubleshooting performance issues

Prerequisites: Link Project First

Important: CLI logs require a linked project with Account Token:

# 1. Set Account Token (NOT project token)
export RAILWAY_API_TOKEN=<your-account-token>

# 2. Link project/environment/service
railway link -p <PROJECT_ID> -e <ENVIRONMENT_ID> -s <SERVICE_ID>

# Example:
railway link -p f8fff4f1-1541-4a47-b509-0c30b9459275 \
             -e 3aa108f4-781d-433c-9eee-875c3dbe903d \
             -s 4b879f5f-5e42-4804-a67c-2a1ff2475cb3

# 3. Verify link
railway status

See cli-commands.md for complete reference.

Operations

Operation 1: View Build Logs

Access build process logs to debug compilation and dependency issues.

CLI Commands:

# Stream build logs (live)
railway logs --build

# Last 100 lines of build logs
railway logs --build --lines 100

# Build logs for specific deployment
railway logs --build <DEPLOYMENT_ID>

# Export build logs to file
railway logs --build --lines 500 > build-logs.txt

Dashboard Access:

1. Navigate to Railway Dashboard → Your Service
2. Click Deployments tab
3. Select deployment
4. View Build Logs section

What to Look For:

• ✅ Dependency installation success
• ✅ Build steps completion
• ⚠️ Warning messages (may indicate future issues)
• ❌ Build failures and error messages
• 📊 Build time (optimization opportunities)

Common Issues:

See Also: railway-troubleshooting for cache busting

Operation 2: View Deploy Logs

Monitor deployment lifecycle and health check status.

CLI Commands:

# Stream deployment logs (live)
railway logs --deployment

# Last N lines of deploy logs
railway logs --deployment --lines 200

# Specific deployment by ID
railway logs --deployment <DEPLOYMENT_ID>

# Export deploy logs
railway logs --deployment --lines 500 > deploy-logs.txt

Dashboard Access:

1. Railway Dashboard → Service → Deployments
2. Select deployment
3. View Deploy Logs section

Deployment Phases:

1. Building - Compiling code
2. Publishing - Creating container image
3. Deploying - Rolling out to infrastructure
4. Health Checking - Verifying service health
5. Active - Deployment live

Health Check Debugging:

# View health check failures
railway logs --deployment | grep "health check"

# Common health check issues:
# - Port not exposed correctly
# - Application not binding to 0.0.0.0
# - Health endpoint not responding
# - Application crashing during startup

Troubleshooting:

• Health check failing? Verify PORT environment variable
• Deployment stuck? Check for blocking startup processes
• Rollback occurring? Check health check configuration

Operation 3: View Runtime Logs

Access application stdout/stderr for debugging runtime behavior.

CLI Commands:

# Stream runtime logs (live, Ctrl+C to stop)
railway logs

# Last N lines (stops streaming)
railway logs --lines 500

# Stream different service/environment
railway logs --service backend --environment production

# Filter with Railway syntax
railway logs --filter "@level:error"
railway logs --lines 100 --filter "timeout"

# Pipe to grep for local filtering
railway logs | grep ERROR

# JSON output for parsing
railway logs --json | jq 'select(.level == "error")'

Dashboard Access:

1. Railway Dashboard → Service → Observability
2. Click Logs tab
3. Real-time log stream with filtering

Structured Logging Best Practices:

Railway supports structured JSON logging. Output JSON on a single line:

// Node.js Example
console.log(JSON.stringify({
  level: 'error',
  message: 'Database connection failed',
  error: err.message,
  timestamp: new Date().toISOString(),
  userId: req.user?.id
}));

# Python Example
import json
import logging

logging.basicConfig(format='%(message)s')
logger = logging.getLogger()

logger.error(json.dumps({
    'level': 'error',
    'message': 'Database connection failed',
    'error': str(e),
    'timestamp': datetime.utcnow().isoformat(),
    'user_id': user_id
}))

Supported Log Levels:

• debug - Detailed diagnostic information
• info - General informational messages
• warn - Warning messages (potential issues)
• error - Error messages (failures)

Benefits:

• All JSON fields are searchable in Railway dashboard
• Better filtering and analysis
• Integration with log aggregation tools

Operation 4: View HTTP Logs

Analyze HTTP request patterns and debug API issues.

Dashboard Access:

1. Railway Dashboard → Service → Observability
2. Click HTTP tab
3. View request metadata

Available Metadata:

• HTTP method (GET, POST, etc.)
• Request path
• Status code
• Response time (ms)
• Client IP address
• User agent
• Timestamp

Filtering HTTP Logs:

# Filter by status code
@httpStatus:500

# Filter by path
@path:"/api/users"

# Combine filters
@httpStatus:500 AND @path:"/api"

Use Cases:

• Identify slow endpoints (high response time)
• Find error patterns (500, 404 status codes)
• Analyze traffic patterns
• Debug API issues
• Monitor rate limits

Performance Analysis:

# Find slow requests (>1000ms)
# Filter in dashboard: responseTime > 1000

# Find all 5xx errors
# Filter: @httpStatus:5xx

# Analyze specific endpoint
# Filter: @path:"/api/checkout"

Operation 5: Filter and Search Logs

Use Railway's powerful filtering syntax for targeted log analysis.

Filter Syntax:

Boolean Operators:

# AND - Both conditions must match
@httpStatus:500 AND @path:"/api"

# OR - Either condition matches
@level:error OR @level:warn

# NOT - Exclude matches
NOT @path:"/health"

# Grouping
(@level:error OR @level:warn) AND @service:api

Common Filter Patterns:

# All errors from last hour
@level:error

# Slow HTTP requests (>1000ms)
@httpStatus:200 AND responseTime > 1000

# Failed API calls
@path:"/api" AND @httpStatus:5xx

# Exclude health checks
NOT @path:"/health" NOT @path:"/metrics"

# Specific user errors
@level:error AND @userId:12345

# Database connection issues
"connection refused" OR "timeout"

Dashboard Filtering:

1. Observability → Logs
2. Enter filter in search box
3. Use dropdowns for common filters
4. Save frequent filters as presets

CLI Filtering:

# Use grep for basic filtering (streaming is default)
railway logs | grep ERROR

# Use jq for JSON logs
railway logs --json | jq 'select(.level == "error")'

# Complex filtering with awk
railway logs | awk '/ERROR/ || /WARN/'

Operation 6: Export Logs Externally

Export logs to external systems for long-term retention and analysis.

Why Export?

• Railway retention: 7-30 days (plan dependent)
• Long-term log storage
• Advanced analytics
• Compliance requirements
• Centralized multi-service logging

External Export Options:

Option 1: Locomotive Sidecar (Webhook Export)

Deploy a sidecar container to forward logs via webhooks.

Repository: https://github.com/railwayapp/locomotive

Setup:

# Add locomotive service to Railway project
railway service create locomotive

# Configure environment variables
WEBHOOK_URL=https://your-log-endpoint.com/ingest
WEBHOOK_METHOD=POST
WEBHOOK_HEADERS='{"Authorization": "Bearer xxx"}'

# Deploy locomotive
railway up

Supported Destinations:

• Custom webhooks
• Datadog
• Axiom
• BetterStack
• Logtail
• Any HTTP endpoint

Option 2: OpenTelemetry (OTEL) Integration

Send logs using OTEL protocol.

Environment Variables:

# Add to your service
OTEL_EXPORTER_OTLP_ENDPOINT=https://otel-collector.example.com:4318
OTEL_EXPORTER_OTLP_HEADERS=x-api-key=xxx
OTEL_SERVICE_NAME=my-railway-service

Supported OTEL Collectors:

• Grafana Alloy
• OpenTelemetry Collector
• Datadog Agent
• New Relic
• Honeycomb

See Also: observability-stack-setup for LGTM stack

Option 3: Log Streaming Script

Use the provided script to stream logs to external systems.

Usage:

# Stream to file
.claude/skills/railway-logs/scripts/stream-logs.sh --output file --path logs/

# Stream to webhook
.claude/skills/railway-logs/scripts/stream-logs.sh --output webhook \
  --url https://logs.example.com/ingest \
  --token YOUR_API_KEY

# Stream to S3
.claude/skills/railway-logs/scripts/stream-logs.sh --output s3 \
  --bucket my-logs-bucket \
  --prefix railway/

Features:

• Continuous streaming
• Automatic reconnection
• Buffering and batching
• Multiple output formats

Option 4: Manual Export

Export logs for ad-hoc analysis.

# Export last 1000 lines
railway logs --lines 1000 > logs-$(date +%Y%m%d-%H%M%S).txt

# Export recent logs (specify number of lines)
railway logs --lines 1000 > logs-recent.txt

# Export and compress
railway logs --lines 5000 | gzip > logs.txt.gz

Scheduled Export (cron):

# Add to crontab (every 6 hours)
0 */6 * * * railway logs --lines 10000 > /backup/railway-logs-$(date +\%Y\%m\%d-\%H\%M).txt

Integration with External Tools

Datadog

# Install Datadog agent in Railway
# Add environment variables:
DD_API_KEY=xxx
DD_SITE=datadoghq.com
DD_LOGS_ENABLED=true
DD_LOGS_CONFIG_CONTAINER_COLLECT_ALL=true

Axiom

# Use locomotive sidecar
WEBHOOK_URL=https://api.axiom.co/v1/datasets/<dataset>/ingest
WEBHOOK_HEADERS='{"Authorization": "Bearer <token>"}'

BetterStack (Logtail)

# Add to your application
LOGTAIL_SOURCE_TOKEN=xxx

# Use Logtail SDK
npm install @logtail/node

Grafana Loki

# Deploy Grafana Agent/Alloy
# Configure to scrape Railway logs
# See observability-stack-setup skill

Best Practices

1. Structured Logging

DO:

console.log(JSON.stringify({ level: 'info', message: 'User login', userId: 123 }));

DON'T:

console.log(`User ${userId} logged in`); // Hard to search/filter

2. Log Levels

Use appropriate severity:

• debug - Development/troubleshooting only
• info - Normal operations
• warn - Potential issues (high memory, slow queries)
• error - Failures requiring attention

3. Sensitive Data

NEVER log:

• Passwords
• API keys
• Credit card numbers
• Personal identifiable information (PII)

Redact sensitive data:

console.log(JSON.stringify({
  level: 'info',
  message: 'User login',
  email: 'u***@example.com', // Redacted
  ip: req.ip
}));

4. Performance

Avoid excessive logging:

// BAD - Logs every request
app.use((req, res, next) => {
  console.log(`Request: ${req.method} ${req.path}`);
  next();
});

// GOOD - Log only errors or important events
app.use((req, res, next) => {
  if (res.statusCode >= 400) {
    console.log(JSON.stringify({ level: 'error', method: req.method, path: req.path, status: res.statusCode }));
  }
  next();
});

5. External Export

For production:

• Export logs to external system (Railway retention is limited)
• Use structured logging (JSON)
• Implement log rotation
• Set up alerts on error patterns

Troubleshooting Common Issues

Quick Reference

Important: CLI vs API

Use CLI for logs, not API. The Railway GraphQL API log queries (deploymentLogs, buildLogs) often return "Problem processing request". The CLI provides reliable log access:

railway logs                    # Runtime logs
railway logs --deployment       # Build/deploy logs
railway logs --lines 100        # Historical logs

See railway-api/references/api-limitations.md for details.

Related Skills

• railway-troubleshooting - Debug deployment issues
• railway-api - Programmatic log access
• observability-stack-setup - LGTM stack with Loki
• railway-project-management - Project setup

References

See the references/ directory for detailed guides:

• cli-commands.md - Complete CLI reference (verified working)
• log-filtering-syntax.md - Complete filter syntax reference
• structured-logging.md - Structured logging best practices
• external-export.md - External export integration guides

Scripts

See the scripts/ directory for automation tools:

• stream-logs.sh - Stream and filter logs
• export-logs.sh - Export logs to file/webhook

Last Updated: 2025-11-26