Backend monitoring

Monitor Vercel deployments, debug production issues, check logs, and understand logging patterns. Use when investigating errors, checking deployment logs, debugging failures, or improving observability.

install
source · Clone the upstream repo
git clone https://github.com/sgcarstrends/sgcarstrends
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/sgcarstrends/sgcarstrends "$T" && mkdir -p ~/.claude/skills && cp -r "$T/.claude/skills/monitoring" ~/.claude/skills/sgcarstrends-backend-monitoring && rm -rf "$T"
manifest: .claude/skills/monitoring/SKILL.md
source content

Monitoring Skill

Monitor Vercel deployments, check logs, and track errors.

Viewing Logs

Vercel Dashboard

  1. Go to Vercel Dashboard → Project → Logs
  2. Filter by:
    • Function logs
    • Build logs
    • Edge function logs
    • Time range

Vercel CLI

# View recent logs
vercel logs <deployment-url>

# Stream logs in real-time
vercel logs <deployment-url> --follow

# Filter by status
vercel logs <deployment-url> --status error

Vercel Analytics

Web Vitals

Monitor in Vercel Dashboard → Analytics:

  • LCP (Largest Contentful Paint)
  • FID (First Input Delay)
  • CLS (Cumulative Layout Shift)
  • TTFB (Time to First Byte)

Speed Insights

Already integrated in the app:

// apps/web/src/app/layout.tsx
import { Analytics } from "@vercel/analytics/next";
import { SpeedInsights } from "@vercel/speed-insights/next";

const RootLayout = ({ children }) => (
  <html>
    <body>
      {children}
      <Analytics />
      <SpeedInsights />
    </body>
  </html>
);

Logging Patterns

The codebase uses standard

console.log/error/warn
for logging:

// Workflow logging pattern
console.log("[WORKFLOW] Posts cache invalidated");
console.log("[COE]", coeResult);

// Error logging pattern
console.error("Error getting post view count:", error);
console.error("Download failed:", { url, status });

Log Prefixes Used:

  • [WORKFLOW]
    - Workflow execution events
  • [COE]
    - COE data processing
  • [CARS]
    - Car data processing

Health Checks

The app has a health check endpoint at

/api/health
:

// apps/web/src/app/api/health/route.ts
export const GET = () => {
  return NextResponse.json(
    { status: "ok", timestamp: new Date().toISOString() },
    {
      status: 200,
      headers: {
        "Cache-Control": "public, s-maxage=60, stale-while-revalidate=300",
      },
    },
  );
};

Test health endpoint:

curl -f https://sgcarstrends.com/api/health || echo "Web unhealthy"

Workflow Error Handling

Workflows use Vercel WDK's built-in error handling:

// apps/web/src/workflows/cars/index.ts
import { FatalError, RetryableError } from "workflow";

// Retryable errors (will be retried automatically)
if (message.includes("429")) {
  throw new RetryableError("AI rate limited", { retryAfter: "1m" });
}

// Fatal errors (will not be retried)
if (message.includes("401") || message.includes("403")) {
  throw new FatalError("AI authentication failed");
}

Error Types:

  • RetryableError
    - Temporary failures (rate limits, network issues)
  • FatalError
    - Permanent failures (auth errors, invalid config)

Debugging Production Issues

1. Check Vercel Logs

# Get deployment URL
vercel ls

# Check logs
vercel logs <url> --follow

2. Check Build Output

  1. Vercel Dashboard → Deployments
  2. Click on deployment
  3. View "Build Logs"

3. Test Endpoint

curl -v https://sgcarstrends.com/api/health

4. Check Function Runtime

Vercel Dashboard → Project → Settings → Functions

  • View runtime version
  • Check memory allocation
  • Review timeout settings

Common Issues

IssueInvestigationSolution
High latencyCheck Vercel Analytics, slow queriesOptimize queries, add caching
Build failuresCheck build logsFix build errors, update dependencies
Function timeoutCheck function logsOptimize code, increase timeout
Cold startsCheck function invocationsConsider edge functions
Workflow failuresCheck Vercel logs for workflow IDReview error type, check retries

Vercel Function Configuration

// API route config
export const config = {
  maxDuration: 60, // seconds (Pro plan allows up to 60s)
};

Cron Job Monitoring

Workflows are triggered by Vercel Cron:

// apps/web/vercel.json
{
  "crons": [
    { "path": "/api/workflows/cars", "schedule": "0 10 * * *" },
    { "path": "/api/workflows/coe", "schedule": "0 10 * * *" },
    { "path": "/api/workflows/deregistrations", "schedule": "0 10 * * *" }
  ]
}

Monitor cron executions:

  1. Vercel Dashboard → Project → Cron Jobs
  2. View execution history and status
  3. Check logs for each execution

Best Practices

  1. Use Log Prefixes: Add
    [CONTEXT]
    prefix for easy filtering
  2. Log Meaningful Data: Include relevant context (month, record counts)
  3. Don't Log Secrets: Never log passwords, tokens, API keys
  4. Use Vercel Analytics: Monitor performance metrics
  5. Check Cron Jobs: Monitor scheduled workflow executions

References