---
name: customerio-load-scale-2d3e3af35b8a
description: |
  Implement Customer.io load testing and scaling.
  Use when preparing for high traffic, load testing,
  or scaling integrations for enterprise workloads.
  Trigger with phrases like "customer.io load test", "customer.io scale",
  "customer.io high volume", "customer.io performance test".
allowed-tools: Read, Write, Edit, Bash(kubectl:*), Bash(curl:*)
version: 1.0.0
license: MIT
author: Jeremy Longshore <jeremy@intentsolutions.io>
---

# Customer.io Load & Scale

## Overview
Load testing and scaling strategies for high-volume Customer.io integrations.

## Prerequisites
- Customer.io integration working
- Load testing tools (k6, Artillery)
- Staging environment with test workspace

## Capacity Planning

### Customer.io Rate Limits
| Endpoint | Limit | Notes |
|----------|-------|-------|
| Track API (identify/track) | 100 req/sec | Per workspace |
| App API (transactional) | 100 req/sec | Per workspace |
| Webhooks (outbound) | Varies | Based on plan |

### Scaling Targets
| Volume | Architecture | Notes |
|--------|--------------|-------|
| < 1M events/day | Single service | Direct API calls |
| 1-10M events/day | Queue-based | Message queue buffer |
| > 10M events/day | Distributed | Multiple workers |

## Instructions

### Step 1: Load Test Script (k6)
```javascript
// load-tests/customerio.js
import http from 'k6/http';
import { check, sleep } from 'k6';
import { Rate, Trend } from 'k6/metrics';

const errorRate = new Rate('errors');
const identifyDuration = new Trend('identify_duration');
const trackDuration = new Trend('track_duration');

const BASE_URL = 'https://track.customer.io/api/v1';
const AUTH = __ENV.CUSTOMERIO_AUTH; // base64(site_id:api_key)

export const options = {
  scenarios: {
    identify_load: {
      executor: 'ramping-rate',
      startRate: 10,
      timeUnit: '1s',
      preAllocatedVUs: 50,
      stages: [
        { target: 50, duration: '1m' },
        { target: 100, duration: '2m' },
        { target: 100, duration: '5m' },
        { target: 0, duration: '1m' },
      ],
      exec: 'identifyScenario',
    },
    track_load: {
      executor: 'ramping-rate',
      startRate: 10,
      timeUnit: '1s',
      preAllocatedVUs: 50,
      stages: [
        { target: 50, duration: '1m' },
        { target: 100, duration: '2m' },
        { target: 100, duration: '5m' },
        { target: 0, duration: '1m' },
      ],
      exec: 'trackScenario',
    },
  },
  thresholds: {
    'errors': ['rate<0.01'],
    'identify_duration': ['p95<500'],
    'track_duration': ['p95<500'],
  },
};

export function identifyScenario() {
  const userId = `load-test-${__VU}-${__ITER}`;
  const payload = JSON.stringify({
    email: `${userId}@loadtest.com`,
    _load_test: true,
    created_at: Math.floor(Date.now() / 1000),
  });

  const start = new Date();
  const res = http.post(
    `${BASE_URL}/customers/${userId}`,
    payload,
    {
      headers: {
        'Authorization': `Basic ${AUTH}`,
        'Content-Type': 'application/json',
      },
    }
  );
  identifyDuration.add(new Date() - start);

  const success = check(res, {
    'identify status is 200': (r) => r.status === 200,
  });
  errorRate.add(!success);

  sleep(0.1);
}

export function trackScenario() {
  const userId = `load-test-${__VU}-${__ITER}`;
  const payload = JSON.stringify({
    name: 'load_test_event',
    data: {
      source: 'k6',
      timestamp: new Date().toISOString(),
    },
  });

  const start = new Date();
  const res = http.post(
    `${BASE_URL}/customers/${userId}/events`,
    payload,
    {
      headers: {
        'Authorization': `Basic ${AUTH}`,
        'Content-Type': 'application/json',
      },
    }
  );
  trackDuration.add(new Date() - start);

  const success = check(res, {
    'track status is 200': (r) => r.status === 200,
  });
  errorRate.add(!success);

  sleep(0.1);
}
```

### Step 2: Horizontal Scaling
```yaml
# k8s/scaled-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: customerio-worker
spec:
  replicas: 3
  selector:
    matchLabels:
      app: customerio-worker
  template:
    metadata:
      labels:
        app: customerio-worker
    spec:
      containers:
        - name: worker
          image: customerio-worker:latest
          resources:
            requests:
              cpu: "500m"
              memory: "256Mi"
            limits:
              cpu: "1000m"
              memory: "512Mi"
          env:
            - name: CONCURRENCY
              value: "10"
---
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: customerio-worker-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: customerio-worker
  minReplicas: 3
  maxReplicas: 20
  metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 70
    - type: External
      external:
        metric:
          name: pubsub.googleapis.com|subscription|num_undelivered_messages
          selector:
            matchLabels:
              resource.labels.subscription_id: customerio-events
        target:
          type: AverageValue
          averageValue: 1000
```

### Step 3: Message Queue Architecture
```typescript
// lib/scaled-processor.ts
import { Kafka, Consumer, EachMessagePayload } from 'kafkajs';
import { TrackClient, RegionUS } from '@customerio/track';

const kafka = new Kafka({
  clientId: 'customerio-worker',
  brokers: process.env.KAFKA_BROKERS!.split(',')
});

const consumer = kafka.consumer({
  groupId: 'customerio-workers',
  sessionTimeout: 30000,
  heartbeatInterval: 3000
});

const client = new TrackClient(
  process.env.CUSTOMERIO_SITE_ID!,
  process.env.CUSTOMERIO_API_KEY!,
  { region: RegionUS }
);

interface CustomerIOEvent {
  type: 'identify' | 'track';
  userId: string;
  payload: any;
}

async function processMessage(message: EachMessagePayload): Promise<void> {
  const event: CustomerIOEvent = JSON.parse(message.message.value!.toString());

  if (event.type === 'identify') {
    await client.identify(event.userId, event.payload);
  } else if (event.type === 'track') {
    await client.track(event.userId, {
      name: event.payload.event,
      data: event.payload.properties
    });
  }
}

async function start(): Promise<void> {
  await consumer.connect();
  await consumer.subscribe({ topic: 'customerio-events', fromBeginning: false });

  await consumer.run({
    partitionsConsumedConcurrently: 10,
    eachMessage: async (payload) => {
      try {
        await processMessage(payload);
      } catch (error) {
        console.error('Processing error:', error);
        // Dead letter or retry logic
      }
    }
  });
}

start().catch(console.error);
```

### Step 4: Rate Limiter for Fair Usage
```typescript
// lib/rate-limiter.ts
import Bottleneck from 'bottleneck';

// Respect Customer.io's 100 req/sec limit
// Leave headroom for other services
const limiter = new Bottleneck({
  reservoir: 80, // 80 tokens
  reservoirRefreshAmount: 80,
  reservoirRefreshInterval: 1000, // per second
  maxConcurrent: 20,
  minTime: 10 // Minimum 10ms between requests
});

// Track rate limit events
limiter.on('depleted', () => {
  console.warn('Rate limiter depleted, requests queued');
});

limiter.on('error', (error) => {
  console.error('Rate limiter error:', error);
});

export async function rateLimitedIdentify(
  client: TrackClient,
  userId: string,
  attributes: Record<string, any>
): Promise<void> {
  return limiter.schedule(() => client.identify(userId, attributes));
}

export async function rateLimitedTrack(
  client: TrackClient,
  userId: string,
  event: string,
  data?: Record<string, any>
): Promise<void> {
  return limiter.schedule(() =>
    client.track(userId, { name: event, data })
  );
}

// Get limiter stats
export function getLimiterStats() {
  return {
    running: limiter.running(),
    queued: limiter.queued(),
    done: limiter.done(),
    reservoir: limiter.reservoir
  };
}
```

### Step 5: Batch Processing
```typescript
// lib/batch-sender.ts
interface BatchConfig {
  maxBatchSize: number;
  maxWaitMs: number;
  concurrency: number;
}

class BatchSender {
  private batch: Array<{ userId: string; operation: 'identify' | 'track'; data: any }> = [];
  private timer: NodeJS.Timer | null = null;
  private processing = false;

  constructor(
    private client: TrackClient,
    private config: BatchConfig = { maxBatchSize: 100, maxWaitMs: 1000, concurrency: 10 }
  ) {}

  add(userId: string, operation: 'identify' | 'track', data: any): void {
    this.batch.push({ userId, operation, data });

    if (this.batch.length >= this.config.maxBatchSize) {
      this.flush();
    } else if (!this.timer) {
      this.timer = setTimeout(() => this.flush(), this.config.maxWaitMs);
    }
  }

  async flush(): Promise<void> {
    if (this.processing || this.batch.length === 0) return;

    if (this.timer) {
      clearTimeout(this.timer);
      this.timer = null;
    }

    this.processing = true;
    const items = this.batch.splice(0, this.config.maxBatchSize);

    // Process in parallel with limited concurrency
    for (let i = 0; i < items.length; i += this.config.concurrency) {
      const chunk = items.slice(i, i + this.config.concurrency);
      await Promise.allSettled(chunk.map(item => this.processItem(item)));
    }

    this.processing = false;
  }

  private async processItem(item: { userId: string; operation: string; data: any }): Promise<void> {
    if (item.operation === 'identify') {
      await this.client.identify(item.userId, item.data);
    } else {
      await this.client.track(item.userId, {
        name: item.data.event,
        data: item.data.properties
      });
    }
  }
}
```

### Step 6: Load Test Execution
```bash
#!/bin/bash
# scripts/run-load-test.sh

# Set credentials
export CUSTOMERIO_AUTH=$(echo -n "$CIO_SITE_ID:$CIO_API_KEY" | base64)

# Run k6 load test
k6 run \
  --out json=results.json \
  --out influxdb=http://localhost:8086/k6 \
  load-tests/customerio.js

# Generate report
k6 run --summary-export=summary.json load-tests/customerio.js

echo "Load test complete. Results in results.json"
```

## Scaling Checklist

- [ ] Rate limits understood
- [ ] Load tests written
- [ ] Horizontal scaling configured
- [ ] Message queue buffering
- [ ] Rate limiting implemented
- [ ] Batch processing enabled
- [ ] Monitoring during tests

## Error Handling
| Issue | Solution |
|-------|----------|
| Rate limited (429) | Reduce concurrency |
| Timeout errors | Increase timeout |
| Queue backlog | Scale workers |

## Resources
- [k6 Documentation](https://k6.io/docs/)
- [Customer.io Rate Limits](https://customer.io/docs/api/track/#section/Limits)

## Next Steps
After load testing, proceed to `customerio-known-pitfalls` for anti-patterns.


---
Source attribution: Dicklesworthstone
Original project: https://github.com/Dicklesworthstone/pi_agent_rust/blob/f28fc46ee0db8e710d95b8b55592cd38d60f2b26/tests/ext_conformance/artifacts/plugins-community/plugins/saas-packs/customerio-pack/skills/customerio-load-scale/SKILL.md
License: MIT
License information: https://opensource.org/license/mit
Original upstream instructions; downloaded from Mega AI Hub.

## Upstream license notice

MIT License (with OpenAI/Anthropic Rider)

Copyright (c) 2026 Jeffrey Emanuel

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

ADDITIONAL RIDER / RESTRICTION (OpenAI / Anthropic):

This rider is part of the "conditions" of this License. In the event of any
conflict between this rider and any other portion of this License, this rider
controls.

"Restricted Parties" means OpenAI, L.L.C.; Anthropic, PBC; any of their
respective Affiliates; and any person or entity acting directly or indirectly
on behalf of, for the benefit of, or under the direction of any of the
foregoing (including any officer, director, employee, contractor, agent,
consultant, service provider, or representative).

Notwithstanding any other provision of this License, no rights are granted to
any Restricted Party. Any purported license, sublicense, assignment, transfer,
or other permission to any Restricted Party is null and void absent the
express prior written permission of Jeffrey Emanuel.

You may not provide, disclose, distribute, sublicense, sell, lease, lend,
host, make available, or otherwise permit access to the Software or any
derivative work of the Software (as defined in applicable copyright law)
(collectively, "Derivative Works") to or for any Restricted Party.

For purposes of this rider, "use" includes, without limitation: copying,
modifying, merging, publishing, distributing, sublicensing, selling,
transferring, making available, hosting, deploying, executing, benchmarking,
testing, analyzing, indexing, or incorporating the Software or any Derivative
Works into any dataset, training corpus, evaluation harness, or pipeline for
machine learning or other automated systems.

This rider applies to the Software and all Derivative Works. As a condition of
use, you agree that this rider is a precondition to exercising any rights
under this License, and you agree that any distribution of the Software or any
Derivative Works must include this rider provision unmodified.

Any breach of this rider automatically and immediately terminates the
permissions granted by this License. Upon termination, you must immediately
cease all use and distribution of the Software and any Derivative Works and
destroy all copies under your control.

You agree that a breach of this rider would cause irreparable harm and that
Jeffrey Emanuel may seek injunctive or other equitable relief to enforce this
rider, in addition to any other remedies available at law. To the maximum
extent permitted by applicable law, the prevailing party in any action to
enforce this rider shall be entitled to recover reasonable attorneys' fees and
costs.

For purposes of this rider, "Affiliate" means any entity that directly or
indirectly controls, is controlled by, or is under common control with the
specified party. "Control" means ownership of more than 50% of the voting
securities or other ownership interest, or the power to direct management or
policies by contract or otherwise.

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.



# Mega AI Hub resource metadata

Creator / publisher: Dicklesworthstone
Original project: https://github.com/Dicklesworthstone/pi_agent_rust/blob/f28fc46ee0db8e710d95b8b55592cd38d60f2b26/tests/ext_conformance/artifacts/plugins-community/plugins/saas-packs/customerio-pack/skills/customerio-load-scale/SKILL.md
Attribution basis: Source repository record

Fields of work: Marketing
Roles: Growth Marketer, Marketing Analyst, Product Marketer
Specialties: API, Testing
Tools: Artillery, K6
Country applicability: Country not specified
Classification basis: Source catalog labels
Instruction SHA-256: 993d67e7ddba92f82b165d342f114bedaffc99d6dea95e731377eb330f1bb6a7