Gestion des limites de débit Clerk

Comprenez et gérez les limites de débit et quotas de l'API Clerk pour éviter la limitation et optimiser votre utilisation. Implémentez des stratégies de retry automatique et de traitement par lot.

Spar Skills Guide Bot
DeveloppementIntermédiaire
22009/03/2026
Claude CodeCursorWindsurfCopilot
#clerk#rate-limiting#api-optimization#typescript#backend

name: clerk-rate-limits description: | Understand and manage Clerk rate limits and quotas. Use when hitting rate limits, optimizing API usage, or planning for high-traffic scenarios. Trigger with phrases like "clerk rate limit", "clerk quota", "clerk API limits", "clerk throttling". allowed-tools: Read, Write, Edit, Grep version: 1.0.0 license: MIT author: Jeremy Longshore jeremy@intentsolutions.io

Clerk Rate Limits

Overview

Understand Clerk's rate limiting system and implement strategies to avoid hitting limits.

Prerequisites

  • Clerk account with API access
  • Understanding of your application's traffic patterns
  • Monitoring/logging infrastructure

Instructions

Step 1: Understand Rate Limits

Clerk API Rate Limits (as of 2024)

| Endpoint Category | Free Tier | Pro Tier | Enterprise | |------------------|-----------|----------|------------| | Authentication | 100/min | 500/min | Custom | | User Management | 100/min | 500/min | Custom | | Session Management | 200/min | 1000/min | Custom | | Webhooks | Unlimited | Unlimited | Unlimited |

Client-Side Limits

  • SDK requests are automatically throttled
  • Browser session: 10 requests/second
  • Token refresh: 1 per 50 seconds (automatic)

Step 2: Implement Rate Limit Handling

// lib/clerk-client.ts
import { clerkClient } from '@clerk/nextjs/server'

interface RateLimitConfig {
  maxRetries: number
  baseDelay: number
}

async function withRateLimitRetry<T>(
  operation: () => Promise<T>,
  config: RateLimitConfig = { maxRetries: 3, baseDelay: 1000 }
): Promise<T> {
  let lastError: Error | null = null

  for (let attempt = 0; attempt < config.maxRetries; attempt++) {
    try {
      return await operation()
    } catch (error: any) {
      lastError = error

      // Check for rate limit error
      if (error.status === 429 || error.code === 'rate_limit_exceeded') {
        const delay = config.baseDelay * Math.pow(2, attempt)
        console.warn(`Rate limited, retrying in ${delay}ms (attempt ${attempt + 1})`)
        await new Promise(resolve => setTimeout(resolve, delay))
        continue
      }

      // Non-rate-limit error, throw immediately
      throw error
    }
  }

  throw lastError
}

// Usage
export async function getUser(userId: string) {
  const client = await clerkClient()
  return withRateLimitRetry(() => client.users.getUser(userId))
}

Step 3: Batch Operations

// lib/clerk-batch.ts
import { clerkClient } from '@clerk/nextjs/server'

// Instead of multiple individual calls
async function getBatchedUsers(userIds: string[]) {
  const client = await clerkClient()

  // Use getUserList with userId filter (single API call)
  const { data: users } = await client.users.getUserList({
    userId: userIds,
    limit: 100
  })

  return users
}

// Paginated fetching with rate limit awareness
async function getAllUsers(batchSize = 100, delayMs = 100) {
  const client = await clerkClient()
  const allUsers = []
  let offset = 0

  while (true) {
    const { data: users, totalCount } = await client.users.getUserList({
      limit: batchSize,
      offset
    })

    allUsers.push(...users)
    offset += batchSize

    if (allUsers.length >= totalCount) break

    // Rate limit friendly delay
    await new Promise(resolve => setTimeout(resolve, delayMs))
  }

  return allUsers
}

Step 4: Caching Strategy

// lib/clerk-cache.ts
import { unstable_cache } from 'next/cache'
import { clerkClient } from '@clerk/nextjs/server'

// Cache user data to reduce API calls
export const getCachedUser = unstable_cache(
  async (userId: string) => {
    const client = await clerkClient()
    return client.users.getUser(userId)
  },
  ['clerk-user'],
  {
    revalidate: 60, // Cache for 60 seconds
    tags: ['clerk-users']
  }
)

// In-memory cache for high-frequency lookups
const userCache = new Map<string, { user: any; timestamp: number }>()
const CACHE_TTL = 30000 // 30 seconds

export async function getUserWithCache(userId: string) {
  const cached = userCache.get(userId)
  if (cached && Date.now() - cached.timestamp < CACHE_TTL) {
    return cached.user
  }

  const client = await clerkClient()
  const user = await client.users.getUser(userId)

  userCache.set(userId, { user, timestamp: Date.now() })
  return user
}

Step 5: Monitor Rate Limit Usage

// lib/clerk-monitor.ts
interface RateLimitMetrics {
  endpoint: string
  remaining: number
  limit: number
  resetAt: Date
}

const metrics: RateLimitMetrics[] = []

export function trackRateLimit(response: Response) {
  const remaining = response.headers.get('x-ratelimit-remaining')
  const limit = response.headers.get('x-ratelimit-limit')
  const reset = response.headers.get('x-ratelimit-reset')

  if (remaining && limit) {
    metrics.push({
      endpoint: response.url,
      remaining: parseInt(remaining),
      limit: parseInt(limit),
      resetAt: reset ? new Date(parseInt(reset) * 1000) : new Date()
    })

    // Alert if approaching limit
    if (parseInt(remaining) < parseInt(limit) * 0.1) {
      console.warn('Approaching rate limit:', {
        remaining,
        limit,
        endpoint: response.url
      })
    }
  }
}

export function getRateLimitMetrics() {
  return metrics.slice(-100) // Last 100 entries
}

Output

  • Rate limit handling with retries
  • Batched API operations
  • Caching implementation
  • Monitoring system

Rate Limit Headers

x-ratelimit-limit: 100
x-ratelimit-remaining: 95
x-ratelimit-reset: 1704067200

Best Practices

  1. Batch requests - Use getUserList instead of multiple getUser calls
  2. Cache aggressively - User data rarely changes in real-time
  3. Use webhooks - Let Clerk push updates instead of polling
  4. Exponential backoff - Retry with increasing delays
  5. Monitor usage - Track rate limit headers

Error Handling

| Error | Cause | Solution | |-------|-------|----------| | 429 Too Many Requests | Rate limit exceeded | Implement backoff, cache more | | quota_exceeded | Monthly quota hit | Upgrade plan or reduce usage | | concurrent_limit | Too many parallel requests | Queue requests |

Resources

Next Steps

Proceed to clerk-security-basics for security best practices.

Skills similaires