Currently canary functions are distinguished by their name and the PR number of the corresponding PR.
This is not enough, as there can be multiple pipeline executions running parallel (at the same time). This can happen if, for example, multiple changes are made within a PR in a short amount of time.
To better and uniquely link google cloud functions to a specific pipeline execution their name should be extended to include a timestamp of the execution start.
The function name may follow this pattern:
<function-name>-<pr-number>-<timestamp>
E.g.:
transform-42-2024-05-13T12:34:56Z
Currently canary functions are distinguished by their name and the PR number of the corresponding PR.
This is not enough, as there can be multiple pipeline executions running parallel (at the same time). This can happen if, for example, multiple changes are made within a PR in a short amount of time.
To better and uniquely link google cloud functions to a specific pipeline execution their name should be extended to include a timestamp of the execution start.
The function name may follow this pattern:
<function-name>-<pr-number>-<timestamp>E.g.:
transform-42-2024-05-13T12:34:56Z