Post-build processing in Bamboo 2.0

Due to Bamboo 2.0's support for distributed builds, the plug-in point for performing post-build processing has been split into two:

CustomBuildProcessor: These plug-in modules are executed on the agent. They have access to the build filesystem, but not to any Bamboo managers.

CustomBuildProcessorServer: These plug-in modules are executed on the Bamboo server, once the build result has been transmitted from the agent. They have access to Bamboo managers, but they are not guaranteed to have access to the build filesystem.

If you have a plug-in that needs to read data from the build filesystem, and make use of Bamboo managers to retrieve and/or store data, you must use the following approach:

On the Bamboo agent:

1. Have a CustomBuildProcessor read the data you require from the build filesystem on the agent;
2. Store the data you require as key-value pairs in the Map<String, String> obtained from calling getCustomData() of the BuildResult;

Then, on the Bamboo server:

1. Have a CustomBuildProcessorServer read your key-value pairs from the BuildResult on the Bamboo server;
2. Make use of the the Bamboo managers to retrieve and store data as required.