Install the following software packages:
- Pyenv installed and
eval "$(pyenv init -)"is added to the shell configuration file. For more details please refer to the PyEnv installation instructions.
- JDK version required to build Elasticsearch. Please refer to the build setup requirements.
- Docker and on Linux additionally docker-compose.
- git 1.9 or better
Check the installation guide for detailed installation instructions for these packages.
Rally does not support Windows and is only actively tested on MacOS and Linux.
Installation Instructions for Development¶
git clone https://github.com/elastic/rally.git cd rally make prereq make install source .venv/bin/activate ./rally
On MacOS Mojave the step
make prereq might fail with the following message:
zipimport.ZipImportError: can't decompress data; zlib not available
This is a known issue with
pyenv. Please see the Github issue for workarounds.
Rally has a built-in auto-update feature when you install it from sources. By default, it will update from the remote named
origin. If you want to auto-update from a different remote, provide
--update-from-remote=YOUR_REMOTE_NAME as first parameter.
To work conveniently with Rally, we suggest that you add the Rally project directory to your
PATH. In case you use a different remote, you should also define aliases in your shell’s config file, e.g.:
alias rally='rally --update-from-remote=elastic ' alias rallyd='rallyd --update-from-remote=elastic '
Then you can invoke Rally or the Rally daemon as usual and have auto-update still work.
Also note that automatic updates are disabled in the following cases:
- There are local (uncommitted) changes in the Rally project directory
- A different branch than
masteris checked out
- You have specified
--skip-updateas the first command line parameter
- You have specified
--offlineas a command line parameter for Rally
Before we can run our first benchmark, we have to configure Rally. Just invoke
./rally configure and Rally will automatically detect that its configuration file is missing and prompt you for some values and write them to
~/.rally/rally.ini. After you’ve configured Rally, it will exit.
For more information see configuration help page.
Key Components of Rally¶
To get a rough understanding of Rally, it makes sense to get to know its key components:
- Race Control: is responsible for proper execution of the race. It sets up all components and acts as a high-level controller.
- Mechanic: can build and prepare a benchmark candidate for the race. It checks out the source, builds Elasticsearch, provisions and starts the cluster.
- Track: is a concrete benchmarking scenario, e.g. the http_logs benchmark. It defines the data set to use.
- Challenge: is the specification on what benchmarks should be run and its configuration (e.g. index, then run a search benchmark with 1000 iterations)
- Car: is a concrete system configuration for a benchmark, e.g. an Elasticsearch single-node cluster with default settings.
- Driver: drives the race, i.e. it is executing the benchmark according to the track specification.
- Reporter: A reporter tells us how the race went (currently only after the fact).
There is a dedicated tutorial on how to add new tracks to Rally.