Cloudbench is a library designed to make Cloud benchmarking easier.
Cloudbench integrates with a web service (not yet publicly available) to upload benchmark results, and compare your Cloud with other offerings.
First, define the configuration file for the benchmark:
[environment]
fio = Path to the fio binary [/usr/local/bin/fio]
pidfile = Path to the pid file to use [var/run/cloudbench.pid]
logfile = Path to the log file to use [/var/log/cloudbench.log]
nobench = Comma-separated list of devices to not use in benchmark []
[general]
size = The size of the file to use in benchmarks [10G]
ramp = The duration of the benchmark ramp time [15 seconds]
duration = The duration of the benchmarks [600 seconds]
[reporting]
endpoint = The endpoint to report data to
username = The username to authenticate with at the endpoint
apikey = The API key to authenticate with at the endpoint
[benchmarks]
blocksizes = Comma-separated list of block sizes to use during benchmarking
depths = Comma-separated list of depths to use during benchmarking
modes = Comma-separated list of modes to use during benchmarkingNext, execute cloudbench and point it to your configuration file
$ cloudbench -c /path/to/config/file.iniBy default, cloudbench looks for its config file in /etc/cloudbench.ini
For simplified ease of use, it's possible to use Cloudbench with Scalr.
Scalr can:
- Provision and attach volumes (on any platform)
- Install, Configure, and Launch Cloudbench
- Bring infrastructure up and down at regular intervals to periodically run benchmarks
See LICENSE.