> What is the overhead on a FUSE filesystem compared to being implemented in the kernel? It depends on your use case. If you serve most of your requests from

> What is the overhead on a FUSE filesystem compared to being implemented in the kernel? It depends on your use case. If you serve most of your requests from kernel caches, then fuse doesn't add any overhead. That was the case for me, when I had a FUSE service running to directly serve all commits from all branches (from all of history) at the same time as directories directly from the data in a .git folder.