Load testing for fairpredictor

[kshitijrajsharma]

Issue is to do load testing fairpredictor in our cloud so that we can compare and be sure how many parallel users we can support

TBD

[kshitijrajsharma]

Load testing first :

Predictor env : dev

AWS fargate : vcpu - 4 , Ram - 8 GB

Parameters :

Web_Concurrency : 4

Total maximum users : 10
No of users / second swampped : 0.1

[kshitijrajsharma]

Scenario 2 :

env : dev
Same device spec (4vcpu , 8 GB )
Web concurrency : 10
Autoscaling condition : 33 % avg CPU

[kshitijrajsharma]

Scenario 3 :

env : Prod
Device spec : 4vcpu , 8 GB
Web concurrency : 10
Autoscaling condition : 30 % avg CPU

Total maximum users : 10
No of users / second swampped : 0.1

[kshitijrajsharma]

Scenario 4 :

env : Prod
Device spec : 4vcpu , 8 GB
Web concurrency : 10
Autoscaling condition : 30 % avg CPU

Total maximum users : 25
No of users / second swampped : 0.1

[kshitijrajsharma]

Scenario 5 :

env : Prod
Device spec : 4vcpu , 8 GB
Web concurrency : 10
Autoscaling condition : 30 % avg CPU

Total maximum users : 50
No of users / second swampped : 0.1

Note : Already 10 machines are running due to previous load testing and we continued the scenario with 50 users

[kshitijrajsharma]

Scenario 6:

env : Prod
Device spec : 4vcpu , 8 GB
Web concurrency : 10
Autoscaling condition : 30 % avg CPU

Total maximum users : 100
No of users / second swampped : 0.1

Note : Already 30 machines are running due to previous load testing and we continued the scenario with 100 users