Hey guys,
I did some simulations with Libp2p gossip pubsub in the context of Ethereum 2.0 block dissemination: https://hackmd.io/ZMBsjqdqSAK026iFFu_2JQ
Would be glad to hear any feedback. Especially from @vyzo and @raul
Hey guys,
I did some simulations with Libp2p gossip pubsub in the context of Ethereum 2.0 block dissemination: https://hackmd.io/ZMBsjqdqSAK026iFFu_2JQ
Would be glad to hear any feedback. Especially from @vyzo and @raul
This is great, thank you.
I would like to add here that we are planning to act upon gossip pretty soon so as to repair the mesh when receiving messages only through gossip. This should bring delivery times down again in an adverse environment.
β vyzo
Another thing I am considering is increasing the default D_lazy
to 12, so that the protocol has higher resistance to this type of sybil attacks.
β vyzo
to repair the mesh when receiving messages only through gossip
@vyzo do you have any PR on this? I could create the appropriate scenario on this and check the effect
increasing the default
D_lazy
to 12, so that the protocol has higher resistance to this type of sybil attacks.
Yep that sounds good
No open PR yet, but itβs part of https://github.com/libp2p/go-libp2p-pubsub/issues/233
β vyzo