Description of problem: Currently if you take a server down from a cluster you can create split-brain issues. Version-Release number of selected component (if applicable): How reproducible: Shut down a server in the middle of an update to a large file to create a split brain issue. Steps to Reproduce: 1. 2. 3. Actual results: You end up with a split brain issue Expected results: The ability to mitigate split-brain issues by putting the server or brick into a maintenance mode that will finish any writes that are being processed and then gracefully shutdown. Additional info:
Migrated to github: https://github.com/gluster/glusterfs/issues/552 Please follow the github issue for further updates on this bug.