I've got a question about the fix for this issue. It looks like it prevents any log messages from being written from an interrupted thread. For example:

try {
  Connection connection = dataSource.getConnection();
} catch (SQLException e) {
  LOGGER.error("Oh no!")
  LOGGER.error("This is bad")
}

In the case where the SQLException is raised as the result of the thread being interrupted while getting the Connection then neither of those messages are written when using logback-classic-1.1.4-SNAPSHOT. The call to blockingQueue.put catches the InterruptedException each time. With versions < 1.1.4-SNAPSHOT the first error message is lost but the second one shows up because the first call cleared the interrupt. It's definitely better to preserve the interrupt flag but halting all logging doesn't seem ideal. Would this be considered a bug?

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira