Debugging space usage discrepancies

Hello bre , elec and other gentlemen!

I am re-writing my post and hope that if it is deleted, it will be done openly and indicating the reason for the deletion.

My node has 45.4 million pieces in the US1 satellite catalog
NodeID: 12epLYjk********
joined 2019-05-31

Currently our largest node has 26M pieces, so bumping the bloom size to 4MiB will probably help…

so… 26 million and 45.4 million.
I already raised the topic about where our money is Lebowski here - Mismatch of proportions, or where is our money, Lebowski? - #16 by elek

As I understand it, it turned out to be an old parable - the path to hell is paved with good intentions. They wanted to save satellite resources - they received unpaid space for operators.

After implementing deletion using the bloom filter, there were discrepancies with the space that these nodes occupy on the disk and the space for which node operators are paid.
Previously, while removing each piece was a separate operation, there were no discrepancies. And the older the node and the more pieces the satellite directory contains, the more pieces that are on the operator’s disk, but there is no payment for this. The most problematic satellite is US1, which has the maximum number of pieces.

The fact that a mistake occurred is normal, it doesn’t happen to anyone. But I propose not to hush it up and delete posts, but to solve it together by holding hands together.

I think it would be nice to compensate operators for all these existing problems in the software, say by 25%

4 Likes