The following article describes the benchmark conducted on Athento at the end of 2022.
The test has been carried out on a virtual server with the following characteristics:
- 120 GB RAM
- 32 vCores (2.3 GHz)
- 400 GB SSD
- 10.000 Mb/s
With the Athento version released in November 2022.
Configuration of the services
PostgreSQL has been configured as recommended by PGTune with:
# DB Version: 12
# OS Type: linux
# DB Type: web
# Total Memory (RAM): 60 GB
# CPUs num: 16
# Connections num: 200
# Data Storage: ssd
max_connections = 200
shared_buffers = 15GB
effective_cache_size = 45GB
maintenance_work_mem = 2GB
checkpoint_completion_target = 0.9
wal_buffers = 16MB
default_statistics_target = 100
random_page_cost = 1.1
effective_io_concurrency = 200
work_mem = 19660kB
min_wal_size = 1GB
max_wal_size = 4GB
max_worker_processes = 16
max_parallel_workers_per_gather = 4
max_parallel_workers = 16
max_parallel_maintenance_workers = 4
Apache 2 has been configured with .....
300 million documents have been created, spread over 30 spaces.
10 document typologies of 20 metadata types each have been created.
1000 users have been created.
User interface performance
Below is the result of the performance test for a scenario with 1000 VUs with a duration of 1 hour and a ramp-up of 60 minutes.
Performance at API level
Below you can see the performance test result for a scenario with 600 VUs with a duration of 1 hour and 30 minutes and a ramp up of 60 minutes.
The results of this test were as follows:
Notice on performance in customized environments
These tests are based on the product's own functionality.
In other environments, times may vary depending on the system load due to customs operations, custom commands, specific configurations, etc.
How can Athento scale?
Given its technology stack, Athento can scale both vertically and horizontally. See articles: