Logging — Animate | AI Wisdom

← Back to topic
🎞 ANIMATE · Mode 3 of 5 · +100 XP— Structured Logging
📋Setup
⚙️Child Logger
🔄Structured
✅Error
🎯 Next: Test →
Step 1 / 4
Speed
JSON Logger Setup
pino outputs JSON — log aggregators (Loki, ELK) can query structured fields💡 Analogy
Like using a spreadsheet instead of a notebook — each row is queryable by any column.
Pino outputs each log as a JSON object. Log aggregators (Grafana Loki, ELK Stack) can then filter by any field — orderId, userId, error type — instantly.
Console Output
> {"level":"info","time":"2024-05-12T18:00:00Z",
>  "service":"order-service","version":"1.5.0",
>  "msg":"Starting up"}
●logger.ts
JSON Logger SetupL1–11
1import pino from 'pino';
2 
3// Structured logger — outputs JSON for log aggregators
4const logger = pino({
5  level: process.env.LOG_LEVEL || 'info',
6  base: { service: 'order-service', version: '1.5.0' },
7  timestamp: pino.stdTimeFunctions.isoTime,
8  serializers: {
9    err: pino.stdSerializers.err,  // safe error serialization
10    req: pino.stdSerializers.req,
11  },
12});
13 
14// Contextual child logger — carries request context automatically
15export function createRequestLogger(requestId: string, userId?: string) {
16  return logger.child({ requestId, userId });
17}
18 
19// Usage: log with structured fields, not string interpolation
20export async function processOrder(orderId: string, req: Request) {
21  const log = createRequestLogger(req.headers['x-request-id'] as string);
22  log.info({ orderId, action: 'started' }, 'Processing order');
23 
24  try {
25    const order = await fetchOrder(orderId);
26    log.info({ orderId, total: order.total, items: order.items.length },
27              'Order fetched successfully');
28 
29    await chargePayment(order);
30    log.info({ orderId, paymentId: order.paymentId }, 'Payment charged');
31 
32  } catch (err) {
33    log.error({ err, orderId }, 'Order processing failed');
34    throw err;
35  }
36}
36 lines
Timeline — 4 Steps
JSON Logger Setup
pino outputs JSON — log aggregators (Loki, ELK) can query structured fields💡 Analogy
Like using a spreadsheet instead of a notebook — each row is queryable by any column.
Pino outputs each log as a JSON object. Log aggregators (Grafana Loki, ELK Stack) can then filter by any field — orderId, userId, error type — instantly.
Console Output
> {"level":"info","time":"2024-05-12T18:00:00Z",
>  "service":"order-service","version":"1.5.0",
>  "msg":"Starting up"}

Was this page helpful?

Discussion

pino outputs JSON — log aggregators (Loki, ELK) can query structured fields

💡 Analogy

Like using a spreadsheet instead of a notebook — each row is queryable by any column.

Pino outputs each log as a JSON object. Log aggregators (Grafana Loki, ELK Stack) can then filter by any field — orderId, userId, error type — instantly.

Console Output

> {"level":"info","time":"2024-05-12T18:00:00Z",
>  "service":"order-service","version":"1.5.0",
>  "msg":"Starting up"}

pino outputs JSON — log aggregators (Loki, ELK) can query structured fields

💡 Analogy

Like using a spreadsheet instead of a notebook — each row is queryable by any column.

Pino outputs each log as a JSON object. Log aggregators (Grafana Loki, ELK Stack) can then filter by any field — orderId, userId, error type — instantly.

Console Output

> {"level":"info","time":"2024-05-12T18:00:00Z",
>  "service":"order-service","version":"1.5.0",
>  "msg":"Starting up"}