If JIRA is forcefully taken down, e.g. power interruption, while updating CVS log, the CVS integration will fail on restart with the message "An unknown error occurred - actions == null." in the Version Control tab of the issue detail page, and with an error in the log. It would be great if the VcsService would clean up locks on startup, and certainly the error message should be improved.

2005-06-23 10:26:49,020 ERROR [atlassian.jira.vcs.DefaultRepositoryManager] Error occurred while updating repository 'ris': Could not obtain lock 'C:\jira\cvs\cvslog.write.lock' in 10000 msecs.
com.atlassian.jira.util.LockException: Could not obtain lock 'C:\jira\cvs\cvslog.write.lock' in 10000 msecs.
at com.atlassian.jira.vcs.cvsimpl.CvsRepositoryUtilImpl.updateCvs(CvsRepositoryUtilImpl.java:246)
at com.atlassian.jira.vcs.cvsimpl.CvsRepository.updateCvs(CvsRepository.java:204)
at com.atlassian.jira.vcs.cvsimpl.CvsRepository.updateRepository(CvsRepository.java:295)
at com.atlassian.jira.vcs.DefaultRepositoryManager.updateRepository(DefaultRepositoryManager.java:489)
at com.atlassian.jira.vcs.DefaultRepositoryManager.updateRepositories(DefaultRepositoryManager.java:445)
at com.atlassian.jira.service.services.vcs.VcsService.run(VcsService.java:54)
at com.atlassian.jira.service.JiraServiceContainer.run(JiraServiceContainer.java:59)
at com.atlassian.jira.service.ServiceRunner.execute(ServiceRunner.java:50)
at org.quartz.core.JobRunShell.run(JobRunShell.java:191)
at org.quartz.simpl.SimpleThreadPool$WorkerThread.run(SimpleThreadPool.java:516)