Cache Fluxnode list

[blondfrogs]

• For users that are managing lots of nodes in the fluxnode.conf, starting them can cause the daemon to lock up and become unresponsive.
• For people calling listfluxnodes all the time, there is now a cache that is retained as long as the block height doesn't change

Fixes for this include:

• Making sure when we loop through vectors of data we are accessing the items by reference(&) instead of copying them into memory.
• Caching the list of available fluxnodes in memory after the first time we loop through all wallet transactions. Instead of processing through every wallet transaction for each node we want to start. I believe this is the biggest contributor to the daemon locking up as we was LOCK and UNLOCK the wallet critical section. This change will at least free up a lot of locks and hopefully will make the processes a lot smooth and produce less errors.

[blondfrogs]

This was tested by a community member running 100+ nodes in the fluxnode.conf, and it solved the issues they were seeing. And increased the speeds by a lot

[Cabecinha84]

Lots of cache talking today eheheh. Is it possible to cache the fluxnode list that os get on rcp, until a new block is processed? We are already using cache on FluxOs side but only for 2 minutes I believe, and with the number of nodes we have currently seems like a heavy call. Cheers

A quinta, 9/03/2023, 23:01, Jeremy Anderson @.***> escreveu:

This was tested by a community member running 100+ nodes in the fluxnode.conf, and it solved the issues they were seeing. And increased the speeds by a lot

— Reply to this email directly, view it on GitHub https://github.com/RunOnFlux/fluxd/pull/217#issuecomment-1462943028, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKEMZLQ3AYTDN2JALYCYI4DW3JOOFANCNFSM6AAAAAAVVSMITA . You are receiving this because your review was requested.Message ID: @.***>

[blondfrogs]

Lots of cache talking today eheheh. Is it possible to cache the fluxnode list that os get on rcp, until a new block is processed? We are already using cache on FluxOs side but only for 2 minutes I believe, and with the number of nodes we have currently seems like a heavy call. Cheers A quinta, 9/03/2023, 23:01, Jeremy Anderson @.> escreveu: … This was tested by a community member running 100+ nodes in the fluxnode.conf, and it solved the issues they were seeing. And increased the speeds by a lot — Reply to this email directly, view it on GitHub <#217 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKEMZLQ3AYTDN2JALYCYI4DW3JOOFANCNFSM6AAAAAAVVSMITA . You are receiving this because your review was requested.Message ID: @.>

Added caching for rpc calls that are listing fluxnodes.