In our CI environment we use TeamCity with SVN and Ant.
We have ~2000 automated acceptance tests which we want to run in parallel on several machines.
We need this in order to get fast feedback, as as step towards Continuous Deployment.
The acceptance tests are using JSystem infrastructure, which basically extends JUnit.
We want to get a bottom line result in a single build configuration for the acceptance tests.
The summary build configuration should display tests details, enabling to display test history and statistics.
It seems that if we use different agent for each machine, we won't have a bottom-line result, and it will be very hard to manage.
Is there a proposed solution for this?