[Bug] [Task] Shell task running error:TaskException: Download resource file: (/shell_test/a.sh,tenant_pydolphin) error

[wmyeason]

Search before asking

• I had searched in the issues and found no similar issues.

What happened

dolphinscheduler version : 3.1.9
Deploy using Docker Composer according to official documentation. In order to enable the resource center to upload files, the common. properties file configuration of the API and worker nodes has been modified according to the document
The following issues have arisen

A new shell task execution script file has been created, with the content of echo '123'

The operation log is shown in the figure

I have already entered the API node to check the existence of this file, and the content is also available. I searched for information online but was unable to find it. I am seeking help from the expert

What you expected to happen

echo 123 success

How to reproduce

new file in resource center ,and add task in workflow to running it

Anything else

No response

Version

3.1.x

Are you willing to submit PR?

• Yes I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct

[SbloodyS]

Using sh shell_test/a.sh relative path instead.

[wmyeason]

@SbloodyS still has problem
i build new b.sh , use relative path

still error

[SbloodyS]

Which storage type did you use?

[wmyeason]

@SbloodyS
use HDFS，According to the official website manual

[SbloodyS]

Can you check if shell_test/a.sh file exists in the hdfs path?

[wmyeason]

@SbloodyS i create new file c.sh and find it in api node ,but still has same error

[SbloodyS]

Looks like your hdfs configuration is not effective. Please check it.

[wmyeason]

@SbloodyS According to the official documentation, after configuring the common. properties file, I am able to upload and create files. Why can it be said that HDFS is not working? How should I check the situation you mentioned

[SbloodyS]

You are using local file storage type which does not supoort cluster deployment.

[wmyeason]

@SbloodyS Is it a problem with using Docker Composer for pseudo cluster deployment? Is this deployment not supported by itself? I don't quite understand what you mean by not supporting