Jonas Bostoen
Jonas Bostoen
Another method that could be introduced and might be clearer in the API is `UnsafeNewTx(inner TxData)`, which does the same as `NewTx(inner TxData)` but without deep copying the `TxData`.
Okay, that makes perfect sense, thank you for clarifying. I'll close this issue!
@mattsse for point 1, this should be fine as the ping checks should go through wether the node is in the table or not. If it wasn't in the table,...
@gakonst I quickly checked how geth does it and it looks like they only run checks for peers in the kademlia DHT as well (same as reth). I'll double check...
If the discovery service is keeping track of the bond duration and performs necessary checks, perhaps it could let the network state know of a failing check through an event?
@mattsse I can help with this. Started to look at high-level implementation strategies and I think it would be a clean SoC if we make `Discovery` responsible for this, because...
Hello @tbenr, any questions we can answer to help with evaluation?
We have implemented explicit peering agreements in Lighthouse as trusted peers: - https://github.com/sigp/lighthouse/pull/5333 If you agree with this approach, we'll open a PR that implements this. They can also be...
I think that makes sense. In the context of Lighthouse they didn't want to introduce another CLI flag, which is why we went with this approach. If you agree, we...
Great writeup, thanks for clarifying all the steps. Also agreed on the strategy! One question: when adding delay / bandwidth to a veth inside of a namespace, is it symmetrical...