Internal Server Error when accessing any board after Epic issue type was deleted

Steps to reproduce:

1. Make sure eveything works fine in the boards.
2. Delete Epic issue type.
3. Add another issue type.
4. Access any board, 'Internal Server Error' message should popup and the detailed stack trace:

   @4000000050ec1a3d0f241b34 2013-01-08 13:08:03,253 TP-Processor12 ERROR sysadmin 788x6973x1 r8tl5y 211.24.247.39,131.103.26.146 /rest/greenhopper/1.0/xboard/config.json [common.error.jersey.ThrowableExceptionMapper] Uncaught exception thrown by REST service
   @4000000050ec1a3d0f242304 java.lang.NullPointerException
   @4000000050ec1a3d0f242304       at com.atlassian.greenhopper.service.issuelink.EpicCustomFieldServiceImpl.getDoneEpicStatusOption(EpicCustomFieldServiceImpl.java:139)
   @4000000050ec1a3d0f2426ec       at com.atlassian.greenhopper.web.rapid.RapidBoardViewModelFactory.getEpicConfig(RapidBoardViewModelFactory.java:296)
   @4000000050ec1a3d0f2432a4       at com.atlassian.greenhopper.web.rapid.RapidBoardViewModelFactory.buildRapidGlobalConfig(RapidBoardViewModelFactory.java:179)
   @4000000050ec1a3d0f24368c       at com.atlassian.greenhopper.web.rapid.RapidBoardViewModelFactory.getRapidViewConfigModel(RapidBoardViewModelFactory.java:166)
   

Workaround:

Since deleting the 'Epic' issue type and then adding another issue type has caused the 'Epic Status' configuration scheme to disappear, so the workaround as follows:

1. At the 'Epic Status' customfield, add a new configuration scheme context (assuming that there is no any context there).
2. Add any options there and make sure you set a default option value.