JMeter Cluster

By using this repository, you can deploy jmeter cluster(a client instance and some server instances) in your aws account. This repository is featured on the iRidge Developer Blog! here!

Usage

Preparation

• First, install dependent packages.

$ pip install pipenv
$ pipenv install
$ pipenv shell

• Second, copy variables yaml file.

$ cp variables.yaml.sample variables.yaml

• Thrid, set values of variables. The mean of each variable is follow;

variable	description
jmeter_region	aws region in which cluster is deployed
jmeter_vpc	vpc id in which cluster is deployed
jmeter_subnet	subnet in which cluster is deployed
jmeter_allowed_ip	souce ip cidr to be allowed to ssh into instances of cluster
jmeter_ami_id	the id of ami for cluster (recommendation: amazon linux 2 ami, restriction: redhat type distribution)
jmeter_client_instance_type	instance type for client
jmeter_server_instance_type	instance type for server
jmeter_key_pair	key pair name of instances of cluster
jmeter_servers_count	the count of servers
jmeter_install_version	jmeter version to be installed
jmeter_heap_initial_size	initial heap memory size (it will be given to Xms)
jmeter_heap_max_size	max heap memory size (it will be given to Xmx)
jmeter_max_meta_space_size	max meta space size (it will be given to XX:MaxMetaspaceSize)
jmeter_max_open_file	max numbers of files to be opened (it will be given to command ulimit -n)

• Forth, set your aws credential. (if needed, aws profile too by exporting env var AWS_PROFILE)

Deploy cluster

• deploy cluster by only executing this command!

$ sceptre --var-file ./variables.yaml launch cluster

Destroy cluster

• destroy cluster by only executing this command...

$ sceptre --var-file ./variables.yaml delete cluster

Tips

Each of JMeter servers has local variables in the following;

variable	description
server_number	the number of each JMeter server
server_count	the total count of JMeter servers

You can use these variables in your test scenarios by describing as ${__P(<variable name>)}.