Hi Team ,
As part of CICD implementation , i need to generate reports such as checkstyle and Javadoc. But I dont want the stage result effect build result. I want the build to pass irrespective of stage or task of report generation succeeds or not . Any suggestion . Tried with final stage inclusion in master but it doesn't resolve my complete requirement . Adding a build fail trigger will not be efficient as it runs all other steps again. Help me here.
Hi @Vineela
Let me comment on your comments. I want to be on the same page as you.
As part of CICD implementation , i need to generate reports such as checkstyle and Javadoc. But I dont want the stage result effect build result.
Can you give me more detais on this part?
"I dont want the stage result effect build result.".
I want the build to pass irrespective of stage or task of report generation succeeds or not . Any suggestion .
We can't configure a build to run restrictively, but if all your stages are configured as final ones, all should run and all the jobs in them should run. The problem is if you have a failed task the next one inside the same job will not run.
Tried with final stage inclusion in master but it doesn't resolve my complete requirement.
What was the requirement missing for you in this case?
Adding a build fail trigger will not be efficient as it runs all other steps again. Help me here.
Can you also describe this part better?
Hi Daniel, the query here is I want a stage to run (lets say reports generation) only if the build is successful . Also I don't want the reports generation stage affect the build result . Just that it should run when build is successful .
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi @Vineela
I think this Atlassian Labs plugin would solve your problem, but it is not updated to the latest version of Bamboo. At the moment of this message, it is compatible with versions between 6.4.0 and 6.7.3 (including them).
Atlassian Labs plugins are not supported and users cannot expect them to be updated or have bugs fixed, but I'm working internally to check if we can update this one considering its utility to multiple customers.
As an alternative, I would try to use a script task as the last task of the last stage to run the reports. I would run in such a way that the error return would be always 0 (which means success). We can do that if we wrap our bash commands with true :
command || true
Another alternative, for now, would be creating a deployment project just to fire the reports and having it configured to run on every successful build.
I hope that helps.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.