Bitcoin full node and query server
Nov 5, 2024
View in context in the libbitcoin Slack #general mirror.
josie18:43
somewhat unrelated, im very curious what insights you guys have regarding mmap'd db architectures. ive been exploring mdbx as a replacement for leveldb in core as sort of a moonshot side experiment, and mdbx (fork of lmdb) is essentially a mmap'd shadow paged btree. we saw some really promising benchmarks, but also a ton of write amplification. curious if you guys have any insights, since it seems you've rolled your own mmaped db architecture?
evoskuil21:00
Similarly fan-out by tx adds additional cost with no gain, if there are sufficient numbers of concurrent blocks to validate such that the CPU is fully occupied. This is almost always the case during initial and catch-up sync, but not at the very start and end of the sync. But still not worth that tiny advantage. On the other hand, it's materially faster to fan-out by tx when validating confirmability by block, given sequential execution by block.