Using docker container default entrypoint

When we use Command Line runner (and I assume all the others that support container wrapper), we can choose Custom Script or Run Executable with Parameters but:

* first replaces image entry point with shell to paste the script
* second has “executable” as mandatory field and that replaces entrypoint as well

I want to use the entrypoint setup in the image without having to re-specify it:

* it's unconvenient to have to look them up if we didn't build the image ourselves
* some entrypoints are long and rather not reuqire them to be maintained/visible in build configurations
* sometimes entrypoints change names while retaining functionality - having to specify them breaks it while using default would keep working


Using Custom Script, the behavior is expected, entrypoint not used.

But when using Run Executable with Parameters, Executable should not be required if a container image name is specified. Or a magic keyword to put there that would mean “nothing for entrypoint, use default”.

I know I can use command line runner WITHOUT image name and simply do docker run MY_IMAGE parameters but then I also have to specify all the pieces that Container Wrapper does (--volume $(pwd):$(pwd) -w $(pwd) --env LIST_OF_ALL_ENV_VARS --etc). And also it looks nicer to leverage the wrapper.

Is there any way to achieve this? Can it be submitted as feature request somewhere if not?


1 comment

Hi, thanks for your feedback. Currently, we can't use a docker image default entrypoint by default. We already have a feature request( here. In the meantime, you can use docker run as a workaround the way you described.


Please sign in to leave a comment.