Central Theme
This video conducts a comprehensive stress test on N8N to determine how different configurations handle heavy workloads. The core question is: How does N8N’s architecture (Single Mode vs. Q Mode) and the underlying cloud hardware (AWS C5.large vs. C5.4xlarge) affect performance, stability, and scalability under pressure?
Key Findings & Arguments
The tests were conducted using K6 load testing software across three demanding scenarios: a single webhook, multiple webhooks, and binary file processing. The results consistently highlight the impact of architecture and hardware.
1. Single Webhook Scenario (Simple Traffic)
- Single Mode: On basic hardware (C5.large), it performs well with up to 100 virtual users but begins to fail under heavier loads.
- Q Mode: On the same basic hardware, Q mode dramatically improves performance, increasing throughput from 15 to 72 requests/second and eliminating all failures at 200 virtual users.
- Hardware Upgrade (C5.4xlarge): In Q mode, a hardware upgrade boosts throughput to over 162 requests/second, demonstrating a 10x improvement over the baseline single-mode setup.
2. Multi-Webhook Scenario (Enterprise Multitasking)
- Single Mode: Performance degrades rapidly, with failure rates reaching 38% under a 200-user load on the smaller server.
- Q Mode: Rescues performance, maintaining a stable 74 requests/second with 0% failures on the same hardware. On the larger server, it achieves 162 requests/second.
3. Binary Data Scenario (Heavy File Uploads)
- Single Mode: Fails significantly even on basic hardware, with a 74% failure rate under heavy load.
- Q Mode: Provides more stability but eventually collapses on smaller hardware. However, on the powerful C5.4xlarge server, Q mode successfully handles the entire test with a 0% failure rate, proving it’s essential for processing large files.
Conclusions & Takeaways
- Q Mode is Not Optional for Scale: It is the single most important factor for improving performance and stability. It decouples request intake from execution, preventing bottlenecks and handling high concurrency effectively.
- Hardware Matters: While Q mode provides the architectural foundation, scaling hardware vertically (more CPU, RAM) is critical for maximizing throughput, especially for multitasking and binary-heavy workflows.
- Plan for Your Workload: Simple trigger-based workflows are light, but multitasking and binary data processing demand robust architecture (Q Mode, multiple workers) and powerful hardware from the start. Don’t wait for your system to fail before you upgrade.
Mentoring Question
Based on your current or planned automations, which of these test scenarios (single webhook, multi-webhook, or binary data) most closely resembles your workload, and what does this tell you about the ideal N8N architecture and hardware you should be using?
Source: https://youtube.com/watch?v=YvOCJzya9wU&si=41nIeZjDWRnokIyJ