Hi everyone! Numic only joined Lido in the most recent onboarding.
During this process we were already made aware that Lido aims to reduce the share of Geth. Consequently, we have decided that our initial Vouch controlled setup will run three EL clients, Nethermind and Erigon, in addition to Geth. This creates more safety and flexibility while keeping us less reliant on the majority client.
Also, as Jim McDonald pointed out on the last call, the problem isn’t necessarily Geth, it’s running a majority client. So if at some point more than half of the validators suddenly start using, say, Nethermind, we could find ourselves in a similar situation again.
At Stakely, the percentage of Geth nodes is 30%, including backup nodes. Our commitment to client diversity is high, as we run Lighthouse, Prysm, Teku, and Nimbus as CLs, and Nethermind, Geth, and Erigon as ELs.
We believe it is not necessary to reduce our Geth setup, as we consider that it contributes to the stability of a diverse client setup like ours, and in case of failures from other clients, it adds reliability. Our goal is for Geth to remain a minority in our setup, staying below 30%.
We will soon be adding Besu again and in the future reth, adding even more diversity.
Software diversity is continuous process in RockX, there are quite some different applications running in our testnet environment, like Besu, Teku, Erigon, Nimbus etc. etc., this is valuable to our team when we need to adopt different options of applications to live in Mainnet.
For NO EL Diversity case, with the different test results given, we committed 20-30% Lido validators to Nethermind by Q1 2024, and it’s already work in progress.
We will keep this diversity further by end of 2024, with more EL CL and validator clients in our Mainnet, not only Lido NO but also other validators managed by RockX.
We’ve always valued client diversity. So we’ve been diversifying our CL clients.
Now it’s EL’s turn. We’re already migrating some Geth to Nethermind. I want to use multiple clients rather than a single client. This approach will ensure that we are not reliant on a single client, thereby enhancing Ethereum network’s robustness. We need to diversify even more.
Hey there!
We as blockscape are commited to switch to a setup w/o Geth for a small set of our validators in Q2. The majority of our validators will continue to use Geth as well as other EL clients for now. If there are no major issues within Q2 we will remove Geth for attestations altogether.