As I was having a lot of errors in sending tx in the before tests I was working in a way to send each tx twice trough two nodes. But unfortunately I couldn't finalize this.
As I could solve and fix the the bugs that caused the previous errors I gave up (at least for now) to work on this. But in the today stress test the tx was sent by a grid of nodes.
This grid was composed of 36 geographically dispersed nodes exclusively to send the txs, that was sent by each one randomically. All interconnected with each other and externally.
We want to have dedicated connection to external nodes also, I will contact you in priv to get the information to connect to your node, this will be very good. Thank you very much!
I have configured my node to maximize connections and set up memory according. I can hold 4GB worth of unconfirmed TX before raising relay fee, and have 8gb for cache out of 32gb.
Hitting the network from all over the world. Great work Esthon! Will TheWildCard be doing analysis of bottlenecks, comparison of implementations, how many nodes dropped out, etc?
No, op_return increases the price of the transactions and doesn't prove anything IMO, anybody can send transactions with the same message, what does it prove?