Bahmni/performance-test

Bahmni Performance Test

Documentation

Local Execution

Simulations can be run locally using Gradle. ./gradlew gatlingRun

Note: By default the simulation would run for Standard traffic. Please set LOAD_SIMULATION_TYPE env variable to High or Peak to simulate other traffic conditions and to dev to run with dynamic load conditions

Pre-requisites

• Ensure at-least 50 patients are active in OPD
• Clone the repository and navigate to root directory

IDE Execution export LOAD_SIMULATION_TYPE='Standard' && ./gradlew gatlingRun
High export LOAD_SIMULATION_TYPE='High' && ./gradlew gatlingRun
Peak export LOAD_SIMULATION_TYPE='Peak' && ./gradlew gatlingRun
dev export LOAD_SIMULATION_TYPE=dev ACTIVE_USERS=40 DURATION=10 UNITS=minutes && ./gradlew gatlingRun

AWS EC2 Execution

• Get the ec2access.pem key from the infra admin
• use the following command to login ssh -i "ec2access.pem" ubuntu@***********.ap-south-1.compute.amazonaws.com
• Copy the scripts 📁 into the EC2 machine
• Run this to start the test nohup bash ./startPerformanceTest.sh dev 40 30 minutes & in the scripts folder
• To view execution progress tail -f nohup.out
• To download report scp -i "ec2access.pem" -r ubuntu@***********.ap-south-1.compute.amazonaws.com:/home/ubuntu/Bahmni/scripts/performance-test/build/reports/gatling/* ./
• To stop the execution kill -9 {JAVA_PID}

Stack

• JDK 11
• Scala 2.13
• Gatling - Async Scala-Akka-Netty based -
• Gradle Wrapper is already packaged, so you don't need to install Gradle (a JDK > 11 must be installed and $JAVA_HOME configured)
• Minimal build.gradle leveraging Gradle wrapper and io.gatling.gradle plugin

Folder layout

📁 api

• All HTTP requests responsible to make API calls to OMRS

📁 configurations

• Possibility : Base object to represent possible variants of various User flows
• Protocol : http protocol for running simulations
• TrafficConfiguration : Contains 4 configurable load configurations such as Standard, High , Peak and dev
• Feeders : csv data Feeders

📁 registries

• Domain object registry that represents various functions performed by the user

📁 scenarios

• Patient count will be calculated based on pace for each user and total duration

• workload will be created for each user based on patient count

• Closed System for simulating constant traffic of concurrent users

// Simulate 10 concurrent users traffic for 5 minutes with initial ramp of 30 seconds
rampConcurrentUsers(0).to(10).during(30 seconds),
constantConcurrentUsers(10).during(5 minutes)

📁 simulations

• Current suit supports Bhamni Clinic simulation
• Following closed system concurrent user configurations are defined in TrafficConfiguration

Load Type	Concurrent Users	Duration	Initial Ramp Duration
STANDARD	40	1 hour	60 seconds
HIGH	50	1 hour	60 seconds
PEAK	70	1 hour	60 seconds
dev	env variable	env variable	10% of Duration

To get an idea how this behaves, have a look at this visual representation of a capacity load test:

Workflow

• Trigger is set as Manual
• The Run Performance Test trigger could be utilised to generate performance test report based CUSTOM_SIMULATION
• The workflow performs the following steps :
  • Starting the EC2 Instance
  • Building the project
  • Running performance test in the EC2 instance based on the LOAD_SIMULATION_TYPE or CUSTOM_SIMULATION
  • To start the custom simulation provide the following as inputs
    • Number of concurrent users to test - Example: 50,70,100
    • Duration of test - Example: 5,10,24
    • Units for duration - Hours or Minutes
    • Valid reportname (optional) - Example: bahmniclinic-20230330090426833
    • Enter number of patients to be created (optional) - Example: 50,100 (These patients will be created and same will be replaced in registrations.csv)
  • By default the startup script which runs before the test closes all the exisiting active patients visits and create 60% of Number of concurrent users as new patients followed by starting the visit for them
  • Follow the steps in AWS EC2 Execution to view the progress and download the report

Note

• Stop Performance Test EC2 instance this workflow can be used to stop the EC2 instance
• Execution in GH Machine this workflow can be used to run the test in Github provided machine which is not recommanded due to hardware restrictions
• Manually executing scrape.py helps to fetch values from gatling reports, according to HEADERS set in .env file, and convert it into a CSV file.
• WIP - To convert CSV files into an interactive dashboard for easier comparison of reports and deducing the best executions based on patient encounter count, no: of concurrent users and so on.