< phantomcircuit>
im a bit confused here "Reject tx code 66: mempool min fee not met:"
< phantomcircuit>
"mempoolminfee": 0.00000000
< phantomcircuit>
what's going on there?
< sipa>
mempoolminfee is 0 when there is no minfee being enforced by the mempool
< sipa>
there may be other reasons why a minimum is needed
< phantomcircuit>
sipa, it's weird im seeing things in debug.log that dont seem to be anywhere in the code base
< phantomcircuit>
OOH im seeing the reject messages from p2p
< phantomcircuit>
ha
< midnightmagic>
does -core remember what tx were dumped and ignore them in the future too (presuming no node restart)
< midnightmagic>
i thought so but now I don't see any tracking mech
< sipa>
dumped as in considered invalid or dumped as in evicted from mempool?
< sipa>
(the answer is yes in both cases)
< sipa>
1) using the rejection cache
< sipa>
2) by increasing the mempool relay fee
< midnightmagic>
sipa: thank you. sorry I missed your response. dumped as in timed-out, or evicted from the mempool via resource limit config. does the relay fee come back down when the mempool empties? (optimistically assuming the spam attacks will eventually stop and go away)
< midnightmagic>
sipa: i'd noticed bluematt's mempool limits were in by default and decided to try them out on a small node i run
< sdaftuar>
hi reviewers -- it would be great to get some feedback/code review on the ancestor fee rate ("child-pays-for-parent") mining code in #7600. i think this code is ready and demonstrably better than the current mining algorithm.
< sipa>
will look at it, once i'm done with some bip9 and segwit updates
< sdaftuar>
great thanks. i'm collecting more data on the performance improvement, it's substantial on recent data