Hi everyone, I have a replicated setup with 3 Gluster nodes and on each I have a brick of the same size. I know that the brick folder should never be written to as changes ar not proragated to replicated gluster setup. It should remain read-only. It has happened on several occasions (manual error and application using wrong path) that the brick folder has been written to. So, I got an inconsistency between the brick and gluster mounted replicated folder. I am using the brick folder to read the files as it is much much faster. My question is what exactly to do in this situation to bring the gluster up to date and resolve the inconsistency? Is it enough to write the file that went written into the brick once more to the gluster mounted read-write path? kind regards -- -- Haris Zukanovic
