Magement Performance Insights

We give you clear insights in what Magement offers your business. In this report we focus on the potential that Magement offers and its possible limitations.

 
performance-monitor-banner.jpg
 

Abstract

The Performance Insights that Magement provides is related to different aspects that we’ll describe in this report. This particular report focuses on an installation of Magement on Google Cloud. Hosting is an important aspect that influences the scalability of the platform.

This report has been written to describe the performance of Magement as a first iteration of performance reports.

 
 

Specifications

 
 
iconfinder_Programming_Development_hardware_1743799 (1).png

Hosting setup

Provider: Google Cloud Platform
Server: 8 virtual processors (vCPU)
Memory: 30GB RAM
Web Server: Apache
PHP version: 7.3

 
iconfinder_Programming_Development_Api_1743800.png
 

Magement setup

Software version: v2.31.0
Components used:
• Webhook API & Subscriber
• Directory Storage Publisher
• Type Setter & Key Filter Transformer

 
 

Architecture

Magement is a Symfony PHP based platform. It uses low-code configuration to connect applications. The modular-based system allows the integrator to apply so-called Transformers to manipulate data in any way needed. Within Magement storages can be setup at any point of a route to optimze the processing of data.

 

 

simple-white-background.png

Test Use Case


 
iconfinder_SEO_site_map_969252.png

Part 1. Subscribing data

Magement has subscribers to import data from a source into Magement. In this test we used Magement Webhooks to offer a subscriber a pool of data, the test set. We iterated subscribing 1,000 records of webhook data 50 times. Every iteration executed as a single thread. The amount of threads can be configured in Magement.

Results:
• Average time taken for 1,000 records: 6.36 seconds
• Median time taken for 1,000 records: 6.32 seconds
• Average entities per second: 157.07 entities

 
 

 
iconfinder_SEO_share_2524418.png

Part 2. Publishing data

Magement uses publishers to export data from Magement to an external destination. In this test we used Magement Queue to offer tasks to a publisher. We iterated the test set 50 times, similar to our approach regarding the subscriber test. Publishers were configured to publish in batches of 1, 10, 20, 50, 100 and 200 tasks at a time in a single thread.

Results:
• Average amount per second: 15.50 tasks
• Median amount per second: 16.92 tasks
• Average time per task: 0.06 seconds
• Median time per task: 0.06 seconds

 
 

 

Reliable platform at peak moments

 
Magement-Performance-Peak-Moments-Ribbons.png
 
 

1,5 million data transfers
per day*

* Based on single-threaded hosting.

iconfinder_SEO_awards_969266 (1).png

100,000 data transfers
per minute*

* Example of multi-threaded hosting.

 
 

Additional performance information:
• Multi-threaded data streams (ex. orders, stock, customers) will perform parallel with comparable performance.
• In this report we focussed on a webhook interface. Using different interfaces, like MysQL or HTTP will provide similar results in performance.

 

 

Interested in a Performance Report on your use case?