Hands-Free Releasing - Top Level View #
The HFR consists of:
- HFR core - Manages the testing runs using information from its peripherals. It processes the input data, generates the actions and performs them.
- Interfaces - Bi-directional communication with peripherals.
The Diagram #
The Core #
The HFR core is responsible for:
- Processing the data from inputs (status of testbeds and testing runs, configuration and operational data)
- Evaluation of current state of testbeds and testing runs
- Generating actions:
- Testing runs - starting, cancelling, checking the status, ….
- Testbeds - checking the status, (un)reservation (next phase)….
- Writing operational data to the storage
- Monitoring
- Availability of operational data
Testbeds and interface to testbeds #
- Uses existing set of testbeds
- Interface
- ssh channel
- bash commands
- Tasks performed on the testbeds (next phase):
- reservation and voiding of testbeds
- TG and DUTs health
GitHub Actions and interface to GHA #
Must be able to:
- start a run with parameters
- check a run status
- cancel a run
Interface
Data storage and interface to data storage #
Configuration of HFR and runs
- GitHub
- the configuration files are part of code
- possibility to store more than one configuration (e.g. rc1, rls, …)
- File system of the executor
- Operational data (current status of HFR and runs).
Managing the testing runs #
HFR does not run any tests, it manages the testing runs and resources (testbeds and GHA runs) used by those testing runs.
The blue box is explained in the chapter The Core
The lifecycle of the testing run #
- Pick the run from the list of runs sorted by priority.
- If there is an available testbed for it, start it.
- The run is queued in GHA, waiting for an executor.
- The run started in GHA and reserves the testbed.
- The run runs tests on the testbed.
- The run finishes testing on the testbed.
- The run releases the testbed.
- The run writes the results of tests (JSON files and logs) to the storage.
- The run finishes.