Flushing the WiredTiger journal immediately with the most up-to-date size storer information when 'fsync()' or 'fsyncLock()' is called on the server would fix a build failure where a node is fsync locked and we copy its file over to a new directory to prevent validation from failing due to a fast count mismatch when starting a new node on the copied files.