Carrier customized Gatling container
getcarrier/perfgun:1.0 - Carrier Perfgun release version 1.0
getcarrier/perfgun:latest - bleeding edge, not recommended for production
These simple steps will run Gatling test against your application.
Example docker invocation:
docker run --rm -t -u 0:0 \
-v <your_local_path_to_tests>:/opt/gatling/user-files/ \
-v <your_local_path_to_config/config.yaml>:/tmp/ \ #optional
-v <your_local_path_ to_reports>:/opt/gatling/results/ \ #optional
-e test=<simulation_name> \
-e "GATLING_TEST_PARAMS=-DapiUrl=<url> -Dduration=60 ..." \ #optional
-e "test_type=<test_type>" \ #optional, default - 'demo'
-e "compile=true" # optional, default - false
-e "env=<env>" \ #optional, default - 'demo'
-e "loki_host={{ http://loki }}" # loki host or IP
-e "loki_port=3100" # optional, default 3100
-e "influxdb_host=<influx_host_DNS_or_IP>" \
getcarrier/perfgun
your_local_path_to_tests - path on your local filesystem where you store gatling tests
simulation_name - name of the gatling simulation (with package name if exist) that will be run
your_local_path_to_reports - path on your local filesystem where you want to store reports from this run
your_local_path_to_config/config.yaml - config.yaml file with InfluxDB, Loki, Jira and Report Portal parameters (described below)
test_type - optional tag, used to filter test results
env - optional tag, used to filter test results
loki_host - loki host or IP, used to report failed requests to Loki
loki_port - optional, default 3100
influxdb_host - InfluxDB host DNS or IP. See InfluxDB configuration below
compile - ( true | false ) flag to either compile sources before execution or not. Compilation increase the time of execution.
If your test requires additional parameters, such as user count, duration, and so on, they can be passed to GATLING_TEST_PARAMS with the -D option.
You can also add Java heap params, like -Xms1g -Xmx1g in the end of the GATLING_TEST_PARAMS
To run demo test you can use this command:
docker run -t -u 0:0 --rm \
-e test=carrier.Flood \
-e "GATLING_TEST_PARAMS=-DapiUrl=https://challengers.flood.io -Dduration=10 -Dramp_users=1 -Dramp_duration=1" \
getcarrier/perfgun:1.0
Report is located in your your_local_path_to_reports folder
#####InfluxDB config
There are 2 ways to transfer parameters for InfluxDB.
The first way is to pass parameters using environment variables (-e "influxdb_host=<influx_host_DNS_or_IP>").
List of all parameters that are used to configure InfluxDB:
influxdb_host - InfluxDB host DNS or IP.
influxdb_port - InfluxDB port. Optional, default - 8086.
influxdb_user - InfluxDB user (login). Optional, default - ''.
influxdb_password - InfluxDB password. Optional, default - ''.
influxdb_database - InfluxDB database where test results will be stored. Optional, default - 'gatling'.
influxdb_comparison - InfluxDB database for the comparison table. Used to compare different tests. Optional, default - 'comparison'.
The second way is to pass the parameters using config.yaml. (Described below)
If you use both options, the values that are passed through the environment variables will be applied.
Reporting can be configured using config.yaml file.
During the test, you can send detailed error messages to Loki. To do this, you need to uncomment the appropriate section in config.yaml and specify host and port for Loki.
In post-processing, the results are analyzed according to three criteria: functional errors, percentage of performance degradation, number of requests exceeding the threshold value.
At the end of the test you can create JIRA tickets or Report Portal launches with the results of post analysis.
To do this, you need to uncomment the necessary configuration section and pass parameters.
config.yaml file example:
# Reporting configuration section (all report types are optional)
#reportportal: # ReportPortal.io specific section
# rp_host: https://rp.com # url to ReportPortal.io deployment
# rp_token: XXXXXXXXXXXXX # ReportPortal authentication token
# rp_project_name: XXXXXX # Name of a Project in ReportPortal to send results to
# rp_launch_name: XXXXXX # Name of a launch in ReportPortal to send results to
# check_functional_errors: False # Perform analysis by functional error. False or True (Default: False)
# check_performance_degradation: False # Perform analysis compared to baseline False or True (Default: False)
# check_missed_thresholds: False # Perform analysis by exceeding thresholds False or True (Default: False)
# performance_degradation_rate: 20 # Minimum performance degradation rate at which to create a launch (Default: 20)
# missed_thresholds_rate: 50 # Minimum missed thresholds rate at which to create a launch (Default: 50)
#jira:
# url: https://jira.com # Url to Jira
# username: some.dude # User to create tickets
# password: password # password to user in Jira
# jira_project: XYZC # Jira project ID
# assignee: some.dude # Jira id of default assignee
# issue_type: Bug # Jira issue type (Default: Bug)
# labels: Performance, perfgun # Comaseparated list of lables for ticket
# watchers: another.dude # Comaseparated list of Jira IDs for watchers
# jira_epic_key: XYZC-123 # Jira epic key (or id)
# check_functional_errors: False # Perform analysis by functional error False or True (Default: False)
# check_performance_degradation: False # Perform analysis compared to baseline False or True (Default: False)
# check_missed_thresholds: False # Perform analysis by exceeding thresholds False or True (Default: False)
# performance_degradation_rate: 20 # Minimum performance degradation rate at which to create a JIRA ticket (Default: 20)
# missed_thresholds_rate: 50 # Minimum missed thresholds rate at which to create a JIRA ticket (Default: 50)
#influx:
# host: carrier_influx # Influx host DNS or IP
# port: 8086 # Influx port
# user: "" # Influx user (login) (Default: "")
# password: "" # Influx password (Default: "")
# influx_db: gatling # Database name for gatling test results (Default: gatling)
# comparison_db: comparison # Database name for comparison builds (Default: comparison)
#loki:
# host: http://loki # Loki host DNS or IP
# port: 3100 # Loki port
Carrier Perfgun can be started inside Jenkins CI/CD pipeline.
Here is an example pipeline that will run demo test.
def get_influx_host(String env_var) {
def match = env_var =~ 'http://(.+)/jenkins'
return match[0][1]
}
node {
stage("configure"){
deleteDir()
sh 'mkdir reports'
sh '''echo """influx:
host: '''+get_influx_host(env.JENKINS_URL)+'''
port: 8086
""" > config.yaml'''
}
stage('execute tests') {
def dockerParamsString = ""
def params = [
"--entrypoint=''",
"-e test=\"carrier.Flood\"",
"-e GATLING_TEST_PARAMS=\"-DapiUrl=https://challengers.flood.io -Dduration=10 -Dramp_users=1 -Dramp_duration=1\""
]
for (param in params) {
dockerParamsString += " ${param}"
}
docker.image("getcarrier/perfgun:1.0").inside(dockerParamsString) {
sh "cp ${WORKSPACE}/config.yaml /tmp/config.yaml"
sh "executor.sh"
sh "cp -R /opt/gatling/results/* ${WORKSPACE}/reports/"
}
}
stage("publish results"){
publishHTML([allowMissing: true, alwaysLinkToLastBuild: false, keepAll: false, reportDir: 'results/*/', reportFiles: 'index.html', reportName: 'Gatling Report', reportTitles: ''])
}
}
In order to run your tests you need to copy your tests or clone your repository with the tests in the Jenkins workspace.
Then, inside the docker.image() block, add the following command:
sh "cp -r ${WORKSPACE}/<path_to_user-files_folder> /opt/gatling"
<path_to_user-files_folder - path in Jenkins workspace where stored Gatling simulation
In our opinion tests on gatling should have 2 different stages - compilation and execution
Precondition for compilation step is bucket availability in object storage
- To create a bucket open a galloper url in the browser e.g.
http://{{ galloper_url }} - Click on Artifacts in the side menu
- Click on the Bucket icon in right side of the page and choose
Create New Bucket - Name your bucket e.g. gatling
Now you can compile your tests.
In order to compile the tests and upload them to object storage please run the following command (with your values for test, galloper_url, bucket and aftifact as well as mount your folder with tests)
Note: artifact should have .zip extension
docker run --rm -t -u 0:0 \
-v <your_local_path_to_tests>:/opt/gatling/user-files/ \
-e artifact="Flood.zip" \
-e galloper_url="http://{{ galloper_url }}" \
-e bucket="gatling" -e compile=true \
getcarrier/perfgun:latest
Once compilation completed you can find it in http://{{ galloper_url }}/artifacts/{{ bucket }} with the name artifact
Run the following command:
docker run --rm -t -u 0:0 \
-e test=carrier.Flood -e galloper_url="http://{{ galloper_url }}" \
-e bucket="gatling" -e artifact="Flood.zip" \
-e "GATLING_TEST_PARAMS=-DapiUrl=https://challengers.flood.io -Dduration=30 -Dramp_users=1 -Dramp_duration=1" \
getcarrier/perfgun:latest
What it will do is copy saved artifact and execute simulation carrier.Flood with parameters from GATLING_TEST_PARAMS