Back to Firecrawl

Scraping Load Testing - Test #2

apps/test-suite/load-test-results/tests-1-5/load-test-2.md

2.9.03.6 KB
Original Source

Scraping Load Testing - Test #2

Summary

The load test encountered significant issues, processing 9000 requests with 5473 timeouts and a 61.6% failure rate. The average response time was 3682.1 ms, with a peak response time of 9919 ms. Both machines reached 100% CPU utilization, leading to severe performance bottlenecks and high failure rates. This indicates the need for substantial optimizations, autoscaling, and further investigation.

Table of Contents

Test environment

Machines

MachineSize/CPU
e286de4f711e86 mia (app)performance-cpu-1x@2048MB
73d8dd909c1189 mia (app)performance-cpu-1x@2048MB

Load #2 - 9000 reqs 7 mins 11 secs (4 phases)

yml

# load-test.yml
- duration: 60
  arrivalRate: 10  # Initial load
- duration: 120
  arrivalRate: 20  # Increased load
- duration: 180
  arrivalRate: 30  # Peak load
- duration: 60
  arrivalRate: 10  # Cool down

Archillery Report

Date: 13:50:08(-0300)

MetricValue
errors.ETIMEDOUT5473
errors.Failed capture or match73
http.codes.2003454
http.codes.40164
http.codes.4029
http.downloaded_bytes0
http.request_rate21/sec
http.requests9000
http.response_time.min929
http.response_time.max9919
http.response_time.mean3682.1
http.response_time.median3395.5
http.response_time.p958024.5
http.response_time.p999607.1
http.responses3527
vusers.completed3454
vusers.created9000
vusers.created_by_name.Scrape a URL9000
vusers.failed5546
vusers.session_length.min1127.6
vusers.session_length.max9982.2
vusers.session_length.mean3730.6
vusers.session_length.median3464.1
vusers.session_length.p957865.6
vusers.session_length.p999607.1

Metrics

Both machines reached 100% CPU utilization, which led to a significant number of request failures (61.6% failure rate).


Conclusions and Next Steps

Conclusions

  1. Performance: The system struggled with 9000 requests, resulting in 5473 timeouts and a mean response time of 3682.1 ms.
  2. CPU Utilization: Both machines experienced 100% CPU utilization, causing severe performance degradation and high failure rates.

Next Steps

Implement an autoscaling solution on Fly.io and conduct tests using the same configurations.