A repository of various ActiveData recipes. A recipe is a small snippet that runs one or more active data queries and returns the output. Queries can sometimes be modified by command line arguments and output can sometimes be post-processed.
Each recipe should try to answer a single question.
pip install active-data-recipes
or
git clone https://github.com/mozilla/active-data-recipes
cd active-data-recipes
python setup.py develop
Run:
adr <recipe> <options>
For a list of recipes:
adr --list
For recipe specific options see:
adr <recipe> -- --help
adr backout_rate [--from <date> [--to <date>]]
Get information on the backout rate on autoland and mozilla-inbound over the given time period.
adr code_coverage --path <path> --rev <rev>
Get code coverage information for the given path at rev. Both arguments are required.
adr files_with_coverage
See how many files in tree have any code coverage at all.
adr inspect
Get information on the ActiveData schema. The above command returns the available tables. To see the columns in a table, run:
adr inspect --table task
adr task_durations
Get information on the longest running tasks. Returns the total count, average runtime and total runtime over a given date range and set of branches.
adr try_efficiency
Prints information on try effifiency. This is a measure of how effective try is at preventing backouts. It is roughly:
1000000 / (total_compute_hours_on_try * backout_rate)
adr try_usage
Prints stats on what percentage of try pushes are being scheduled with various different mechanisms
over the last week. The date range can be modified the same as the hours_on_try recipe.
adr try_users
Prints stats on how often individual users are pushing to try over the last week. The date range can
be modified the same as the hours_on_try recipe.