Help Running a local docker image without the use of biocontainers

I’m trying to run a local docker image on galaxy but i’m running into issues.

I’m trying to be able to user the container type=“docker” approach to run a script within the docker container (run.sh).

The galaxy user is able to run docker commands and ran a docker run command within the tool. But I had to mount the input and output files to files within the tool

I was able to create a tool definition that ran docker run --rm -v ${input_file}:/data/file -v ${output_file}:output test_tool -i file -k ${arg} -o outfile

Below is an example of my current tool definition

I get the error that run.sh is not found (the run.sh script is in the entry point to the container and I’ve also given it an absolute path to the script in the command with the same error). It isn’t actually launching the docker container as the documentation would suggest that it should since I define it as a docker container.

Am I missing some configuration parameters in one of the files?

What do I need to do to be able to point a tool at a local docker container

<tool id="docker_tool" name="runs the docker image" version="1.0">
    <description> For a given protein fasta </description>
    <command> <![CDATA[
        run.sh --FASTA ${input_file} --arg1 ${arg} --OUTFILE ${outfile}
        ]]> </command>
    <inputs>
        <param format="fasta" name="input_file" type="data" label="Source file" />
        <param name="arg" type="select" label="Get" multiple="false" display="checkboxes">
            <option value="foo">Foo</option>
            <option value="bar">Bar</option>
            <option value="baz">Baz</option>
        </param>
    </inputs>
    <outputs>
        <data format="tabular" name="outfile" />
    </outputs>
    <container id="test_tool" type="docker">
        <docker>
            <param name="docker_cmd">docker</param>
            <param name="docker_image">test_tool</param> 
        </docker>
    </container>
    <help>
Filler help info 
    </help>
</tool>