Configuration

The vedatrace function accepts a configuration object with the following options:

import { vedatrace, RuntimeType } from 'vedatrace'

const logger = vedatrace({
  apiKey: 'your-api-key',           // Required for HTTP transport
  service: 'my-service',            // Default service name
  endpoint: 'https://ingest.vedatrace.dev/v1/logs',
  environment: 'production',
  
  // Runtime
  autoStart: false,                 // Manual timer control
  runtime: 'cloudflare' as RuntimeType,  // Force runtime
  
  // Batching
  batchSize: 100,                   // Flush after N logs
  flushInterval: 5000,             // Flush after N ms
  
  // Flush behavior
  immediateFlush: true,             // Flush immediately on each log
  waitUntil: waitUntil,             // Cloudflare Workers waitUntil function
  
  // Retry
  maxRetries: 3,
  retryDelay: 1000,
  
  // Redaction
  redaction: {
    paths: ['password', 'token'],
    mask: '[REDACTED]'
  },
  
  // Debug
  debug: true,                      // Enable verbose logging
  
  // Callbacks
  onError: (err) => console.error('VedaTrace error:', err),
  onSuccess: () => console.log('Logs sent')
})

Configuration Options

• apiKey (string, required): Your VedaTrace API key.
• service (string): A default name for your service or application.
• endpoint (string): The VedaTrace ingestion endpoint. Defaults to https://ingest.vedatrace.dev/v1/logs.
• environment (string): The environment your application is running in, e.g., production, development, staging.
• batchSize (number): The number of logs to batch before flushing. Defaults to 100.
• flushInterval (number): The maximum time in milliseconds to wait before flushing a batch. Defaults to 5000 (Node.js / Browser) or 1000 (edge runtimes).
• immediateFlush (boolean): When true, flushes logs to the transport immediately when a log is added, bypassing batching. Useful for serverless environments where timers aren't reliable.
• waitUntil (function): The waitUntil function from cloudflare:workers. When provided, VedaTrace uses it to keep the worker alive during log flushing. This is the recommended approach for Cloudflare Workers.
• maxRetries (number): The maximum number of times to retry sending a batch. Defaults to 3.
• retryDelay (number): The delay in milliseconds between retries. Defaults to 1000.
• redaction (object): An object with paths (an array of strings to redact) and mask (the string to replace them with).
• debug (boolean): When true, enables verbose logging of SDK internals (flush timing, transport errors, batch operations) to the console.
• onError (function): A callback function that is called when an error occurs while sending logs.
• onSuccess (function): A callback function that is called when logs are sent successfully.
• autoStart (boolean): Automatically start the flush timer when the logger is created. Defaults to false for edge runtimes (Cloudflare Workers, Deno, Bun) and true for Node.js and browser.
• runtime (RuntimeType): Force a specific runtime environment. Options: 'node', 'browser', 'cloudflare', 'deno', 'bun', 'edge'. Defaults to auto-detected runtime.