[Improvement] [sub_process] the subprocess's tenant is always default and cann't run

[tuoluzhe8521]

[Improvement] [sub_process] the subprocess's tenant is always default and cann't run

[bajiaolong]

This issue can be avoided by manually selecting tenants. So I think this is an improvement rather than a bug. Anyone who wants to fix it can leave a message under this issue, and I will assign it to it.

I found that I enabled it in the project preferences, but it didn't take effect.

I started the process without setting it according to my project preferences,I don't know if it's my project or if everyone has this problem.

If everyone has this problem, can we improve it like this.
The settings for the tenant are as follows:

• 1. If a project preference is set, the priority is the highest.
• 2. Then the tenant that currently starts the user is the second priority.
• 3. The last priority is the default tenant.

Other configurations are set according to project preferences, and default values are used if project preferences are not enabled.

[ruanwenjun]

Does this happen in the spark task? Isn't the spark task executed by hadoop user?
I just test sub_process task, this work well.

[bajiaolong]

Does this happen in the spark task? Isn't the spark task executed by hadoop user? I just test sub_process task, this work well.

I have encountered a problem and need to modify this tenant every time. the default tenant is default.

1. There is only one tenant in my tenant management: appuser;
   

2. When the process is to run, this tenant must be set up every time, changing the default tenant default to the tenant appuser in my system.
   
   

[tuoluzhe8521]

Does this happen in the spark task? Isn't the spark task executed by hadoop user? I just test sub_process task, this work well.
my task is shell，the shell task call the spark task and the spark task executed by hadoop user.
how can i solve this problem?

[tuoluzhe8521]

Does this happen in the spark task? Isn't the spark task executed by hadoop user? I just test sub_process task, this work well.
this my main task page:

this my sub_process page:

[tuoluzhe8521]

Does this happen in the spark task? Isn't the spark task executed by hadoop user? I just test sub_process task, this work well.

I have encountered a problem and need to modify this tenant every time. the default tenant is default.

1. There is only one tenant in my tenant management: appuser;
   
2. When the process is to run, this tenant must be set up every time, changing the default tenant default to the tenant appuser in my system.
   
   

i delete the default tenant, and also encounter this problem

[Anubisxcw]

Does this happen in the spark task? Isn't the spark task executed by hadoop user? I just test sub_process task, this work well.

I have encountered a problem and need to modify this tenant every time. the default tenant is default.

1. There is only one tenant in my tenant management: appuser;
   
2. When the process is to run, this tenant must be set up every time, changing the default tenant default to the tenant appuser in my system.
   
   

same as u,and i found in lsat version ds ,process defination table has tenant,so you do not nedd to change tenant every time,bug 3.2.1 cancle this coulme in table.terrible

[Anubisxcw]

just delete the default tenant plz.nobody use it in product env,it's useless @rickchengx @SbloodyS

[tuoluzhe8521]

Do you have time to help me take a look @ruanwenjun

[SbloodyS]

This issue can be avoided by manually selecting tenants. So I think this is an improvement rather than a bug. Anyone who wants to fix it can leave a message under this issue, and I will assign it to it.

[Anubisxcw]

This issue can be avoided by manually selecting tenants. So I think this is an improvement rather than a bug. Anyone who wants to fix it can leave a message under this issue, and I will assign it to it.

the point is you need manual set it every time ,because the fronted is stack in 'default' tenant

[lucaswanlinz]

@bajiaolong Just FYI, I encountered the same problem and resolved it by adding an execution environment for the task and configuring "export HADOOP_USER_NAME=your spark user".