execute next step lie on last step's result

Hi ,

Ihave a project which includes 3 steps, there has a scenario that if the step 2 pass, step 3 will run automatically, else step 3 does not run.

How to implement that?

Hi,

That type of flow logic is usually handled by assigning a pass/fail chain on the step. You could create a library or project with the contents of step 3 and assign it as a pass chain on step 2.

Brent Ulbricht
RTC Build Lead

Hi ,

Ihave a project which includes 3 steps, there has a scenario that if the step 2 pass, step 3 will run automatically, else step 3 does not run.

How to implement that?

Hi,

That type of flow logic is usually handled by assigning a pass/fail chain on the step. You could create a library or project with the contents of step 3 and assign it as a pass chain on step 2.

Brent Ulbricht
RTC Build Lead

Thanks!
The contents of step3 is about setting some env such as ".bset env "v1=aaaaaaaaaaaaaaaa"", I try to move to another project and go through the process that you advise, the env can not effect my project.
Maybe the .bset is only effective in steps of a project.
Thoughts?

Hi ,

Ihave a project which includes 3 steps, there has a scenario that if the step 2 pass, step 3 will run automatically, else step 3 does not run.

How to implement that?

Hi,

That type of flow logic is usually handled by assigning a pass/fail chain on the step. You could create a library or project with the contents of step 3 and assign it as a pass chain on step 2.

Brent Ulbricht
RTC Build Lead

Thanks!
The contents of step3 is about setting some env such as ".bset env "v1=aaaaaaaaaaaaaaaa"", I try to move to another project and go through the process that you advise, the env can not effect my project.
Maybe the .bset is only effective in steps of a project.
Thoughts?

Hi,

I know there have been some other topics in this forum about this. The issue is that the pass/fail chain is a separate job and the those variables are separate from the launching project. You could use the .set dot command in the pass chain project and then you will need to have a step in the launching project which explicitly sets that environment group again so that the new variable value is updated.

Brent Ulbricht
RTC Build Lead

There is another way around this problem I've found useful. If the variable is defined in the main job, either in the environment or on the fly, and it is NOT defined in the onpass or onfail project, it is accessible in the called job.

Easy illustration project with one step as follows:

.bset env "testvar=blah"


Create a second project, also with one step:

echo $testvar

set the second project to run onPass of first, and you will see it echoes the variable. This will not work if you also try to define $testvar in the called project's environment.

Hope this helps,
Jonas Gryder