We have many very complex automated deployment tasks which get started using "Send Files or execute commands over SSH", calling some bash shell scripts on one target system connecting to other systems (database, loadbalancer...).
We use the "Editable Email Notification" to get informed about success and failure of the builds/deployments.
We must use the option "Fail the build if an error occurs" for the SSH remote access, in order to catch up all errors.
In certain condition, when exectly one text message (echo) gets printed when the build fails, we know for sure it is no deployment/admin failure but definately an application/development issue.
If the message occurs during "successfull" builds, the developrs can get an automated email, informing them about the not successfully starting up application.
If the message occurs during a build that "failed" due to some application startup issues, the developers do not get immediately informed about the issue, and the system administrators have to dig into the logs, just to find out deployment failed due to some regression.
What would solve this for us:
It would be very nice to have an additional option (separate checkbox, unchecked by default to retain current plugin behaviour) to change even a "failed" build to "unstable" if a specific text pattern occurs in the logs.
EDIT: Spelling corrected