Skip to content

Latest commit

 

History

History
58 lines (44 loc) · 2.5 KB

getting-set-up.md

File metadata and controls

58 lines (44 loc) · 2.5 KB

Getting started with the Performance Dashboard

Prerequisites

  1. Make sure you have Python 2.7.x installed. Python 3.x is not supported.
  2. Download the Google Cloud SDK.
  3. Update the Cloud SDK and set the default project to your project ID by invoking the following commands:
    gcloud components update
    gcloud components install app-engine-python
    gcloud config set project [PROJECT-ID]
    
    Replace [PROJECT-ID] with your project ID. For chromeperf.appspot.com, it's chromeperf.

Running the tests

To run the Python unit tests, use bin/run_py_tests. To run the front-end component tests, use bin/run_dev_server_tests.

Running a local instance

Run bin/dev_server; this sets up a temporary directory, adds links to required libraries, and calls dev_appserver.py on that directory. By default, this starts a server on localhost:8080.

To load sample graph or alert data from production, navigate to /load_from_prod.

Deploying to production

To deploy, you can run bin/deploy, which prepares the code to be deployed and runs gcloud app deploy. If you modify any *.yaml files, you can pass them as parameters to bin/deploy to deploy the updated configs.

When deploying services, bin/deploy doesn't set the new version as the default version; to do this, you can use the Versions page on the Google Developers Console if you have edit or owner permissions for the App Engine project; otherwise if you want to request to set a new default version for chromeperf.appspot.com you can contact [email protected].

After deploying, there is a checklist to verify that no major functionality has regressed: deploy checklist.

WARNING: Some changes to production may not be easily reversible; for example appcfg.py ... vacuum_indexes will remove datastore indexes that are not in your local index.yaml file, which may take more than 24 hours, and will disable any queries that depend on those indexes.

Where to find documentation