Shortage of the current implementation of Soft Updates
The current FreeBSD Soft Updates poses to a performance peanlity when there’re plenty of i-nodes and this is very common case in badly-designed systems, however, I think this should be solved.
According to Kirk’s paper, this can be prevented by committing the snapshot without having to write it synchoriously.
However, I don’t have a filesystem sized 1TB and a small block size. A UFS2? How can I simulate this situation?