KubernetesScriptRunner issues spotted during QA about kestra HOT 7 CLOSED

I can check but as long as logs are sent to DefaultLogConsumer it should be good 🤔

from kestra.

@anna-geller the outputs are fixed in kestra-io/plugin-kubernetes#99

from kestra.

file outputs are not filtered out. When specified to capture CSV files, we get all files that were copied into a working directory, including all namespace files.

This is the case for all existing runner, I agree it's a strange behavior and am willing to reconsider it but this would be a breaking change I think. If you think it would be good to only make available files listed inside output files, which I agree please open a separate issue so we do it for all runners and flag it as a breaking change.

Some properties are marked as required even though they have default values. It seems they should be marked optional

They are required for the script runner to work but as we provide a default value, if the user didn't set any it works. But the user can specify a null value in this case the validation will fail.

from kestra.

@brian-mulier-p FYI the Kuernetes runner dind't create outputs from log, can you check the AWS Batch for it also?

from kestra.

flag it as a breaking change

why would it be a BC for a Beta feature we are only now starting to build? please explain

If we provide default values, it would be intuitive to make it an optional property as it can be skipped then by the user; could you change that?

from kestra.

why would it be a BC for a Beta feature we are only now starting to build? please explain

Because this is not specific to the Kubernetes script runner, the existing Process and Docker runners do the same

from kestra.

@anna-geller

Some properties are marked as required even though they have default values. It seems they should be marked optional

As this one is present everywhere, I close it as won't fix ;)

from kestra.