cloudstash-characterization

Please read Readme.md in each folder for documentation

Running Benchmarks

Use the scripts run_sequential_benchmark.sh and run_load_test_benchmark.sh to orchestrate running the benchmark, this will create a docker container, that will first deploy a cloudstash instance, then run the benchmark, output the results to .csv saved in the output directory, and finally remove the created cloudstash instance.

>>> THE SCRIPTS EXPECTS TO BE CALLED FROM THE ROOT OF THE REPOSITORY <<<

You may add the following option flags after the required arguments:

• --local runt the script locally, will bindmount the benchmark code from the local directory to make development easier
• --tail will automatically follow the generated log file, note that when invoking the option, when you exit tail with Ctrl-c, you will the container process too, so do not use in production!
• --debug will enable some debug prints
• --verbose will enable more verbose prints

Sequential Uploads Benchmark

This benchmark will sequentially upload the specified number of artifacts.

You must pass the number of artifacts to be sequentially uploaded as the first argument:

bash run_sequential_benchmarks <number of artifacts (int)> [options]

Load Test Benchmark

This benchmark will run a predefined load test, consisting of listing artifacts, getting (downloading) artifacts and uploading artifacts.

You must pass the number of artifacts to already be in the cloudstash instance before the load test is started, these will be uploaded to the deployed cloudstash instance before the load test starts.

bash run_load_test_benchmark.sh <number of artifacts (int)> [options]