inspectTree with checksum option - many files open at once when they shouldn't be
metabench opened this issue · 2 comments
A few minutes ago I started an inspectTree call, getting sha256 checksums, on a multi-GB file path. While I expected it to be slow, I did not expect it to open a VERY large number of files at once.
It would be nice to be able to set the number of parallel file reads, or have a reasonably low default such as 4.
Possibly outside the scope of this project, I suggest using an Observable to return tree results as they come in, or at least having that option.
Thanks for reporting. Will see in a free time how I implemented this (didn't remember from the top of my head).
Fixed in v4.0.1
I've just hardcoded arbitrary number of 5 maximum concurrent reads. This is easy to parametrize, but I didn't see such necessity now.