[Bug] When [Module Name] contains multiple sub-processes, the task instance in the sub-process instance obtains duplicate parameters of the pre-task instance.

[AlanZQ]

Search before asking

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

What happened

在工作流中，创建多个子流程，例如创建12个子流程，子流程在获取上游任务参数时，会出现覆盖/错误/以及12个节点拿到的参数完全一致，这不符合预期
通过数据库中查询来看，生成的12的子流程实例，参数没有问题，但是在varpool的设置阶段，获取parent_process_instance_id所对应的varpool，是相同的，这导致子流程实例中的任务实例，执行时的参数永远是相同的

What you expected to happen

通过数据库中查询来看，生成的12的子流程实例，参数没有问题，但是在varpool的设置阶段，获取parent_process_instance_id所对应的varpool，是相同的，这导致子流程实例中的任务实例，执行时的参数永远是相同的
通过debug setSubProcessParam这个方法，能够看出每个子流程任务实例获取到的varpool参数都有问题，并且会随机出现重复

How to reproduce

我已经验证了3.2.1的standalone镜像，3.2.2的standalone镜像，debug镜像代码，都存在此问题，也可以在这两个镜像web服务页面，复现。我提供了我的验证工作流json，⚠️注意：导入后需要更改主流程中子流程的引用为新的id
workflow_1733366901095.json

Anything else

No response

Version

3.2.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

[github-actions]

Search before asking

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

What happened

In the workflow, create multiple sub-processes, for example, create 12 sub-processes. When the sub-process obtains the parameters of the upstream task, there will be coverage/error/ and the parameters obtained by the 12 nodes are exactly the same, which is not as expected.
Judging from the query in the database, there is no problem with the parameters of the generated 12 sub-process instances. However, in the varpool setting phase, the varpool corresponding to the parent_process_instance_id is obtained, which results in the task instance in the sub-process instance being executed. Parameters are always the same

What you expected to happen

Judging from the query in the database, there is no problem with the parameters of the generated 12 sub-process instances. However, in the varpool setting phase, the varpool corresponding to the parent_process_instance_id is obtained, which results in the task instance in the sub-process instance being executed. Parameters are always the same
Through the debug setSubProcessParam method, it can be seen that there is a problem with the varpool parameters obtained by each sub-process task instance, and duplicates occur randomly.

How to reproduce

I have verified that this problem exists in the standalone image of 3.2.1, the standalone image of 3.2.2, and the debug image code. It can also be reproduced on the two image web service pages. I provided my verification workflow json, ⚠️Note: After importing, you need to change the reference of the sub-process in the main process to the new id
workflow_1733366901095.json

Anything else

No response

Version

3.2.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]

Please using english to describe.

[AlanZQ]

Please using english to describe.

In the workflow, create multiple sub-processes, for example, create 12 sub-processes. When the sub-process obtains the parameters of the upstream task, there will be coverage/error/ and the parameters obtained by the 12 nodes are exactly the same, which is not as expected.
Judging from the query in the database, there is no problem with the parameters of the generated 12 sub-process instances. However, in the varpool setting phase, the varpool corresponding to the parent_process_instance_id is obtained, which results in the task instance in the sub-process instance being executed. Parameters are always the same

I have verified that this problem exists in the standalone image of 3.2.1, the standalone image of 3.2.2, and the debug image code. It can also be reproduced on the two image web service pages. I provided my verification workflow json, ⚠️Note: After importing, you need to change the reference of the sub-process in the main process to the new id

[SbloodyS]

Can you try dev to see if this issue exists?

[AlanZQ]

Can you try dev to see if this issue exists?

sure, off course

[AlanZQ]

Can you try dev to see if this issue exists?

ok, I will try later

[AlanZQ]

Can you try dev to see if this issue exists?

Encountered an error logging into the dev image.

It is already the latest version.

Index digest
sha256:6b9c76ee279fb25be83521af5d7b60c19a285672f6963b52eb483698fb90d55b

[ruanwenjun]

Is this issue fixed? I will reopen this issue.

[AlanZQ]

Is this issue fixed? I will reopen this issue.
No, I have verified that both version 3.2.2 and the latest version have this issue. The dev version cannot log in correctly and will throw an error.

[ruanwenjun]

Is this issue fixed? I will reopen this issue.
No, I have verified that both version 3.2.2 and the latest version have this issue. The dev version cannot log in correctly and will throw an error.

I reopen this issue. BTW, this is a known problem, when this feature designed, I have pointed this problem, right now all varpool are stored at the workflow instance, this means once multiple varpools has the same key, then no one can know the value, and only one value will be overwrite.