Open xtc0r opened 1 year ago
Hey! Thanks for the feedback. This has been answered in Discord though will reference it here as well:
We are actively considering and planning client diversity as a significant milestone for 2024. First, we need to get metrics up - similar to
clientdiversity.org
, as we can not solve a problem we can't measure. We're also in contact with 3rd-parties like staking services and Dappnode, so they run minority clients or provide new default images.These steps should already level out the used client majorities, without creating huge problems for private operators. However, articles and awareness regarding this topic are definitely on our list too.
I think we should not wait and fight for something like clientdiversity.org. Turning the tide will only get more difficult because validators will not change their clients in production. Steering the people into the right direction when setting up their clients is much easier! And we also have statistics. See the screenshot of my consensus client below. We have about 83% prysm validators. This value is quite representative as I'm connected to 84 clients and we only have some hundred nodes.
For execution client the geth share is even higher.
@fhildeb Hey Felix, any updates on your comment about something like clientdiversity.org
? Any idea when it will be available? Could I help somewhere?
There is no update for client diversity analysis, as we have limited resources. It would help us if the community or group of network operators could investigate and analyze data sources- then we could create or fork a frontend like clientdiversity.org
and dock on own metrics.
CL Analytics Resources:
EL Analytics Resources:
Research:
What needs to be found out:
Type
Feature
Description
Currently we have with Geth and Prysm clients above 33% majority, with Geth leading the execution clients by far. This can lead to major issues for the chain when Geth or prysm have a critical bug.
TODOs