“Measure and benchmark reliability of your OpenStack virtual machines.”

“VM Reliability Tester” is a software that tests performance and reliability of virtual machines that are hosted in an OpenStack cloud platform. It evaluates the failure rate of VMs by performing a stress test on them. VM Reliability Tester installs OpenStack virtual machines, uploads a test program to them, runs this test program remotely and then captures program execution times to determine reliability of the virtual machines. If the test program takes a significant amount of time to complete, this is considered to result in a VM failure. Such deviations in execution time are an important benchmark for testing performance and reliability of your OpenStack environment.

Clone the git repository into the directory of your choice:

git clone https://github.com/icclab/vm-reliability-tester

Edit the following files:

1. config.ini: The configuration of VM Reliability Tester on your desktop.

2. remote_config.ini: The configuration of VM Reliability Tester on the master VM

3. openrc.py: Your OpenStack Authentication credentials

4. vm_list.csv: Names of your master and client VMs

Run:

python vm_reliability_tester.py

VM Reliability Tester does the folling:

• First it creates the master and client VMs.

• Then it uploads all configuration files and programs to the master VM

• It uploads all test scripts from the master VM to client VMs

• It starts the test runner on the master VM

• The test runner orchestrates the test runs on the client VMs

• The tests are executed on the client VMs

• Measurements are downloaded from the client VMs to the master VM

• The master VM consolidates measurements in the file f_rates.csv

• f_rates.csv is downloaded to your labtop

• Models are created and fitted to the data and stored in fitted_models.csv

• A second remote stress test is initiated

• The master VM orchestrates the 2nd stress test

• The 2nd stress test is executed on the client VMs

• Measurements are downloaded from the client VMs to the master VM

• The master VM consolidates measurements in the file f_rates.csv

• The fitted models are tested versus the 2nd sample and stored in validated_models.csv

• The result is an estimation of failure rates in f_rates.csv and a model of failure rates in validated_models.csv

After the test (it takes 20-30 minutes!) check the files f_rates.csv, fitted_models.csv and validated_models.csv on your labtop.