Skip to main content

Timeouts

Use case​

Abort requests that take too long, either at the client level or per request.

Smallest working example​

import { createClient } from '@parcely/core'

const http = createClient({
  baseURL: 'https://api.example.com',
  timeout: 5000, // 5 seconds for all requests
})

// Override per request
await http.get('/slow-endpoint', { timeout: 30000 })

How it works​

parcely creates an AbortSignal.timeout(ms) and combines it with any user-provided signal via AbortSignal.any. When the timeout fires, the request throws HttpError with code: 'ERR_TIMEOUT'.

import { isHttpError } from '@parcely/core'

try {
  await http.get('/slow')
} catch (err) {
  if (isHttpError(err) && err.code === 'ERR_TIMEOUT') {
    console.log('Request timed out')
  }
}

Axios equivalent​

// axios:
await http.get('/slow', { timeout: 5000 })

// parcely (identical):
await http.get('/slow', { timeout: 5000 })

In axios, the timeout error has code: 'ECONNABORTED'. In parcely it is code: 'ERR_TIMEOUT'.

Notes and gotchas​

  • Timeout is in milliseconds.
  • Timers are cleaned up in finally -- no leaked timers on success or other errors.
  • If both timeout and signal are provided, whichever fires first aborts the request. A user signal abort produces ERR_ABORTED; a timeout produces ERR_TIMEOUT.
  • A timeout of 0 or undefined means no timeout.