Skills Development Handling API Rate Limits with Backoff

Handling API Rate Limits with Backoff

v20260423
flexport-rate-limits
This module provides robust patterns for managing rate limits and throttling when integrating with external APIs (e.g., Flexport). It covers three key strategies: monitoring rate limit headers (X-RateLimit-*), implementing exponential backoff with jitter for transient 429 or 5xx errors, and utilizing queue-based throttling for safe, controlled bulk operations. It ensures resilient and reliable API communication.
Get Skill
499 downloads
Overview

Flexport Rate Limits

Overview

The Flexport API v2 enforces rate limits per API key. When exceeded, you get a 429 Too Many Requests with Retry-After and X-RateLimit-* headers. Key limits to know: the API returns headers on every response telling you remaining quota.

Rate Limit Headers

HeaderDescriptionExample
X-RateLimit-LimitMax requests per window100
X-RateLimit-RemainingRemaining in current window47
X-RateLimit-ResetUnix timestamp when window resets1711234567
Retry-AfterSeconds to wait (only on 429)30

Instructions

Step 1: Monitor Rate Limit Headers

class RateLimitTracker {
  remaining = Infinity;
  resetAt = 0;

  update(headers: Headers) {
    this.remaining = parseInt(headers.get('X-RateLimit-Remaining') || '100');
    this.resetAt = parseInt(headers.get('X-RateLimit-Reset') || '0') * 1000;
  }

  async waitIfNeeded() {
    if (this.remaining <= 2 && Date.now() < this.resetAt) {
      const wait = this.resetAt - Date.now() + 100;
      console.log(`Rate limit near. Waiting ${wait}ms`);
      await new Promise(r => setTimeout(r, wait));
    }
  }
}

Step 2: Exponential Backoff with Jitter

async function flexportWithRetry<T>(
  fn: () => Promise<Response>,
  maxRetries = 4
): Promise<T> {
  for (let attempt = 0; attempt <= maxRetries; attempt++) {
    const res = await fn();

    if (res.ok) return res.json();

    if (res.status === 429) {
      const retryAfter = parseInt(res.headers.get('Retry-After') || '60');
      const jitter = Math.random() * 2000;
      const delay = retryAfter * 1000 + jitter;
      console.log(`429 rate limited. Retry in ${(delay / 1000).toFixed(1)}s`);
      await new Promise(r => setTimeout(r, delay));
      continue;
    }

    if (res.status >= 500 && attempt < maxRetries) {
      const delay = Math.pow(2, attempt) * 1000 + Math.random() * 1000;
      await new Promise(r => setTimeout(r, delay));
      continue;
    }

    throw new Error(`Flexport ${res.status}: ${await res.text()}`);
  }
  throw new Error('Max retries exceeded');
}

Step 3: Queue-Based Throttling

import PQueue from 'p-queue';

// Limit to 10 requests per second with max 3 concurrent
const flexportQueue = new PQueue({
  concurrency: 3,
  interval: 1000,
  intervalCap: 10,
});

async function throttledRequest(path: string): Promise<any> {
  return flexportQueue.add(() =>
    fetch(`https://api.flexport.com${path}`, {
      headers: {
        'Authorization': `Bearer ${process.env.FLEXPORT_API_KEY}`,
        'Flexport-Version': '2',
      },
    }).then(r => r.json())
  );
}

// Bulk operations stay within limits
const shipmentIds = ['shp_001', 'shp_002', 'shp_003', /* ... */];
const results = await Promise.all(
  shipmentIds.map(id => throttledRequest(`/shipments/${id}`))
);

Error Handling

ScenarioStrategy
Single 429Honor Retry-After header
Repeated 429sIncrease backoff, reduce concurrency
Bulk importUse p-queue with intervalCap
Batch readsPaginate with per=100 to minimize calls

Resources

Next Steps

For security configuration, see flexport-security-basics.

Info
Category Development
Name flexport-rate-limits
Version v20260423
Size 3.77KB
Updated At 2026-07-13
Language