The aim of the first section was to introduce the basics concepts behind metadata file systems, to clarify how it differs from regular file systems, and what the advantages and disadvantages are. All of this are good as an introduction, but we have not covered any details yet. As non production ready implementation of a meta file system exits, it would not make any sense to dig in to details about how certain implementation is working. Instead of doing that I will discuss how a metadata file system could be designed and implemented, which problems you will face if doing so, and walk through some of choices which would have to make.