Kai Krakow

Bees will break existing reflinks if it sees other duplicate chains of blocks matching the ones just found. But it will eventually clean up the unreachable extents after some time,...

If your dataset in snapshots is already highly deduplicated, you can try stopping bees, open `beescrawl.dat` and set the `min_transid` to the value of `max_transid` per each line. Then it...

> Just tested `scripts/beesd` (into which I had to hardcode the path to the rebuilt `bees` executable) and it seems to be ok. Is that what you meant? I don't...

Thanks for the explanation, I'll try that. And thanks for all the work that goes into bees. Every little step counts. :-)

Okay, I stopped bees, opened `vim` and did: ```vim :%s/min_transid \(\d\+\) max_transid \(\d\+\)/min_transid \2 max_transid \2/g ``` (after reading your reply again, this is probably not exactly what wanted me...

Looks like it completed a cycle. It then started (according to toxic logs) to read huge amounts of data from my rw subvols (I installed some updates, downloaded some stuff),...

Okay, then extent splitting probably explains what I am seeing. Also, it now reads at varying throughput from 5 MB/s to 800 MB/s, so it actually looks like it is...

Maybe degrade these (and similar) from warning to notice (or info) because there's an active workaround in bees? ```cpp BEESLOGWARN("WORKAROUND: abandoned toxic match for hash "

@Zygo, thank you. All looks good now. Looks like all the split extents have been cleaned up.

I ran into this again because my system froze, and after a reset, xfs cleared the contents of a partially written `beescrawl.dat`. Not only is this especially harmful for performance...