b74ea2e4f8
This adds a new group of commands : 'rally use' the first implemented is 'rally use deployment <deploy_id>' As it is not possible in python to export globaly an env var,it saves the variable in a file at ~/.rally/deployment like : RALLY_DEPLOYMENT=1234abcd For each command that takes a deploy_id as parameter, this parameter is now optional. The algorithm to find the deploy_id is as follows : - If the --deploy-id param is set, use it - If the environ contains the key 'RALLY_DEPLOYMENT', use it - load the content of ~/.rally/deployment and if it contains the key RALLY_DEPLOYMENT, use it - Else, raise an InvalidArgumentException Note that I have not been able to use the default arg in @cliutils.args because it is called before the effective method call. Setting deploy_id=envutils.default_deployment_id() has the same problem Partially implements blueprint: rally-use-command Change-Id: I1e62e26d259bd9e81f1c008058ee9ce91cc3290c
Rally
Introduction
Rally is a Benchmark-as-a-Service project for OpenStack.
Rally is intended to provide the community with a benchmarking tool that is capable of performing specific, complicated and reproducible test cases on real deployment scenarios.
In the OpenStack ecosystem there are currently several tools that are helpful in carrying out the benchmarking process for an OpenStack deployment. To name a few, there are DevStack and FUEL, which are intended for deploying and managing OpenStack clouds, the Tempest testing framework which validates OpenStack APIs, some tracing facilities like Tomograph with Zipkin. The challenge, however, is to compile all these tools together on a reproducible basis. That can be a rather difficult task since the number of compute nodes in a practical deployment can easily be large and also because one may be willing to use many different deployment strategies that pursue different goals (e.g., while benchmarking the Nova Scheduler, one usually does not care of virtualization details, but is more concerned with the infrastructure topologies; while in other specific cases it may be the virtualization technology that matters). What Rally aims to do is Compile many already existing benchmarking facilities into one project, making it flexible to user requirements and ensuring the reproducibility of test results.
Architecture
Rally is split into 4 main components:
Deployment Engine, which is responsible for processing and deploying VM images (using DevStack or FUEL according to user’s preferences). The engine can do one of the following:
- deploy an Operating System (OS) on already existing VMs;
- starting VMs from a VM image with pre-installed OS and OpenStack;
- delpoying multiple VMs inside each OpenStack compute node based on a VM image.
VM Provider, which interacts with cloud provider-specific interfaces to load and destroy VM images;
Benchmarking Tool, which carries out the benchmarking process in several stages:
- runs Tempest tests, reduced to 5-minute length (to save the usually expensive computing time);
- runs the user-defined test scenarios (using the Rally testing framework);
- collects all the test results and processes the by Zipkin tracer;
- puts together a benchmarking report and stores it on the machine Rally was lauched on.
Orchestrator, which is the central component of the system. It uses the Deployment Engine, to run control and compute nodes, in addition to launching an OpenStack distribution. After that, it calls the Benchmarking Tool to start the benchmarking process.
Links
Wiki page:
Launchpad page:
Code is hosted on github: