-
-
AppSignal
Monitoring that respects your time & budget. APM, error tracking, and dashboards for modern web apps. Ten-minute setup, transparent flat pricing, and support from engineers who actually use the product.
-
-
Something like that exists for btrfs; it's called bdtu. It has the accuracy/time trade-off you're interested in, but the implementation is quite different. It samples random points on the disk and finds out what file path they belong to. The longer it runs the more accurate it gets. The readme is good at explaining why this approach makes sense for btrfs and what its limitations are.
https://github.com/CyberShadow/btdu
-
Did you consider the fts[0] family of functions for traversal? I use that along with a work queue for filtered entries to get pretty good performance with dedup[1].
0 - https://linux.die.net/man/3/fts_read
1 - https://github.com/ttkb-oss/dedup/blob/6a906db5a940df71deb4f...
-
- my work display flamegraph in terminal: https://github.com/laixintao/flameshow
-
> For the FAM, your example looks like it requires a compile-time constant size. That's the same as hardcoding an array size in the struct, defeating the whole point. Short names will waste space, and long ones will get truncated.
You made me realize that the const generics stabilization work hasn't advanced enough to do what I was proposing (at least not in as straightforward way): https://play.rust-lang.org/?version=nightly&mode=debug&editi...
Those are const arguments, not const values, which means that you can operate on values as if they didn't have a size, while the compiler does keep track of the size throughout.
