Imported from jeremylongshore/tons-of-skills-marketplace (
skills/.curated/vercel-observability/SKILL.md). Install upstream withnpx skills add jeremylongshore/tons-of-skills-marketplace --skill vercel-observability. Copyright stays with the author (MIT).
Vercel Observability
Overview
Configure comprehensive observability for Vercel deployments using built-in analytics, runtime logs, log drains to external providers, OpenTelemetry integration, and custom instrumentation. Covers the full observability stack from function-level metrics to end-user experience monitoring.
Prerequisites
- Vercel Pro or Enterprise plan (for log drains and extended retention)
- External logging provider (Datadog, Axiom, Sentry) — optional
- OpenTelemetry SDK — optional
Instructions
Step 1: Enable Vercel Analytics
In the Vercel dashboard:
- Go to Analytics tab
- Enable Web Analytics (Core Web Vitals, page views)
- Enable Speed Insights (real user performance data)
// For Next.js — add the analytics component
// src/app/layout.tsx
import { Analytics } from '@vercel/analytics/react';
import { SpeedInsights } from '@vercel/speed-insights/next';
export default function RootLayout({ children }) {
return (
<html>
<body>
{children}
<Analytics />
<SpeedInsights />
</body>
</html>
);
}Install: npm install @vercel/analytics @vercel/speed-insights
Step 2: Runtime Logs
# View runtime logs via CLI
vercel logs https://my-app.vercel.app --follow
# Filter by level
vercel logs https://my-app.vercel.app --level=error
# View logs via API
curl -s -H "Authorization: Bearer $VERCEL_TOKEN" \
"https://api.vercel.com/v2/deployments/dpl_xxx/events?limit=50&direction=backward" \
| jq '.[] | {timestamp: .created, level: .level, message: .text}'Runtime logs include:
- Function invocation start/end with duration
console.log/warn/erroroutput from functions- Edge Middleware execution logs
- HTTP request/response metadata
Step 3: Structured Logging in Functions
// lib/logger.ts — structured JSON logging
interface LogEntry {
level: 'info' | 'warn' | 'error';
message: string;
requestId?: string;
duration?: number;
[key: string]: unknown;
}
export function log(entry: LogEntry): void {
// Vercel captures console output as runtime logs
const output = JSON.stringify({
...entry,
timestamp: new Date().toISOString(),
region: process.env.VERCEL_REGION,
env: process.env.VERCEL_ENV,
});
switch (entry.level) {
case 'error': console.error(output); break;
case 'warn': console.warn(output); break;
default: console.log(output);
}
}
// Usage in API route:
export async function GET(request: Request) {
const requestId = crypto.randomUUID();
const start = Date.now();
try {
const data = await fetchData();
log({ level: 'info', message: 'Fetched data', requestId, duration: Date.now() - start });
return Response.json(data);
} catch (error) {
log({ level: 'error', message: 'Data fetch failed', requestId, error: String(error) });
return Response.json({ error: 'Internal error', requestId }, { status: 500 });
}
}Step 4: Log Drains (External Providers)
Configure log drains to send all Vercel logs to your logging provider:
In dashboard: Settings > Log Drains > Add
Supported providers:
| Provider | Type | Setup |
|---|---|---|
| Datadog | HTTP | API key + site URL |
| Axiom | HTTP | API token + dataset |
| Sentry | HTTP | DSN |
| Custom | HTTP/NDJSON | Any HTTPS endpoint |
| Grafana Loki | HTTP | Push URL + auth |
Log drain delivers:
- Runtime logs: function invocations, console output
- Build logs: build step output, warnings, errors
- Static logs: CDN access logs (edge)
- Firewall logs: WAF events
# Create a log drain via API
curl -X POST "https://api.vercel.com/v2/integrations/log-drains" \
-H "Authorization: Bearer $VERCEL_TOKEN" \
-H "Content-Type: application/json" \
-d '{
"name": "my-datadog-drain",
"type": "json",
"url": "https://http-intake.logs.datadoghq.com/api/v2/logs",
"headers": {"DD-API-KEY": "your-datadog-api-key"},
"sources": ["lambda", "edge", "build", "static"]
}'Step 5: OpenTelemetry Integration
// instrumentation.ts (Next.js 13.4+)
import { NodeSDK } from '@opentelemetry/sdk-node';
import { OTLPTraceExporter } from '@opentelemetry/exporter-trace-otlp-http';
import { getNodeAutoInstrumentations } from '@opentelemetry/auto-instrumentations-node';
export function register() {
const sdk = new NodeSDK({
traceExporter: new OTLPTraceExporter({
url: process.env.OTEL_EXPORTER_OTLP_ENDPOINT,
}),
instrumentations: [getNodeAutoInstrumentations()],
serviceName: 'my-vercel-app',
});
sdk.start();
}// next.config.js
module.exports = {
experimental: {
instrumentationHook: true,
},
};Step 6: Error Tracking with Sentry
npx @sentry/wizard@latest -i nextjs// sentry.client.config.ts
import * as Sentry from '@sentry/nextjs';
Sentry.init({
dsn: process.env.NEXT_PUBLIC_SENTRY_DSN,
environment: process.env.VERCEL_ENV,
release: process.env.VERCEL_GIT_COMMIT_SHA,
tracesSampleRate: process.env.VERCEL_ENV === 'production' ? 0.1 : 1.0,
});Monitoring Dashboard Checklist
| Metric | Source | Alert Threshold |
|---|---|---|
| Error rate | Runtime logs | > 1% of requests |
| P95 function latency | Vercel Analytics | > 2s |
| Cold start frequency | Runtime logs | > 20% of invocations |
| Build success rate | Build logs | Any failure |
| Core Web Vitals (LCP) | Speed Insights | > 2.5s |
| Edge cache hit rate | Static logs | < 80% |
Output
- Vercel Analytics and Speed Insights enabled
- Structured JSON logging in all functions
- Log drains configured to external provider
- Error tracking with Sentry or equivalent
- OpenTelemetry tracing for distributed systems
Error Handling
| Error | Cause | Solution |
|---|---|---|
| Logs missing | Log retention expired (1hr free, 30d with Plus) | Enable log drains for persistence |
| Analytics not tracking | Missing <Analytics /> component |
Add to root layout |
| Log drain not receiving | Wrong URL or auth headers | Test the endpoint directly with curl |
| Sentry not capturing errors | DSN not set in production env | Add NEXT_PUBLIC_SENTRY_DSN to Production scope |
| OTEL traces missing | instrumentation.ts not loaded | Enable instrumentationHook in next.config.js |
Examples
Add trace correlation without retaining user content
Generate a request correlation ID at the edge, propagate it through server-side spans, and log route, duration, status, and a redacted error classification only. Send logs to the approved drain with least-privilege credentials and a documented retention period, then test alerting with a synthetic failure in preview. Before production rollout, confirm the dashboard can link a trace to the deployment SHA while excluding tokens, cookies, and request bodies.
Resources
- Vercel Observability
- Runtime Logs
- Vercel Analytics
- Speed Insights
- Log Drains
- OpenTelemetry + Next.js
Next Steps
For incident response, see vercel-incident-runbook.