Examples of rank CLI usage
All flags in the following commands should be optional, and the CLI should prompt the user for any missing information, allowing them to select from the available options where appropriate.
List all the indices in the rank cluster.
Create a new index in the rank cluster, using data from a copy of a production index (also in the rank cluster), and a locally defined settings/mapping config.
Update the settings/mapping config for an existing index in the rank cluster. Should allow the user to run the update-by-query for all documents immediately, or leave it to be run later.
Delete an index in the rank cluster.
Fetch a copy of the settings/mapping config for an index in the rank cluster.
Copy an index from the rank cluster to a production cluster using cross-cluster replication.
Check the progress/status of a reindex or update-by-query task.
Cancel a task.
Run a search against an index in the rank cluster, outputting formatted results to stdout.
Fetch a set of real search terms from the reporting cluster for each content type, and write them to a local file.
n
should specify the number of terms to fetch, and timestamp
should specify the point in time before which search terms should be retrieved. If a timestamp
is not provided, the current time should be used. Specifying a timestamp allows users to fully reproduce a test at a later date or in a different environment.
Run searches against an index in the rank cluster using a set of real search terms, and compare the speed of the candidate query against the production query.
Run a test against an index in the rank cluster. Test ID should be optional, and if not provided, all tests should be run.
List the available tests using a modified pytest --collect-only
Last updated