Following the EF announcement about pivoting to known-safe hash functions like SHA or BLAKE, leanVM was rewritten to use flock. The new implementation, merged into the leanVM repo today, brings both security and performance to Lean Ethereum.
Our discv5 PR just landed in ethlambda. Lean nodes can now find peers on their own instead of being handed a static bootnode list.
Off by default, you can enable it with --discovery.enable.
We ran a 1024-validator devnet to push the limit on devnet size and aggregation throughput. We are quite happy with the result.
We share our findings on this blogpost.
New blog post out: what we learned running a 1024-validator devnet
In this post, we go through what running our largest Lean Ethereum devnet so far taught us about where the bottlenecks are, and how we can move forward.
While a big contribution to the future of Ethereum, ethlambda is only a piece of what Lambda contributes to the ecosystem.
In this post, we write about what we worked on and what we're building, including ethlambda, @ethrex_client, and Lambda VM.
New blog post about our contribution to the core of Ethereum is out!
Go check out the development status of @ethrex_client, @ethlambda_lean, and Lambda VM.
Last week we fixed a bug in our pruning logic that was quietly costing us CPU and time to finality. The shape of the graph tells the story: things get worse when pruning starts, and go back to normal once the fix lands.
ALT Graphs showing Finalization and Justification delay on the top, and CPU usage on the bottom. All graphs show an increase from the left baseline after pruning was enabled, and a decrease back to the baseline once the fix was deployed.
We're running a 512-validator devnet with ethlambda, as another step towards scaling to higher validator counts.
32 nodes, 4 subnets, one aggregator per subnet, 16 validators per node.
New blog post out: ethlambda, how we made our blocks arrive on time
In this post, we talk about how slow the proposer really is, and show how it can be made fast enough, without changing the underlying cryptographic primitives.
Lean Ethereum client teams are running interop devnets, hunting for any bugs that show up. As well, teams are testing client optimizations to squeeze as much aggregation throughput as possible from leanVM.
To make it more transparent, we made our central devnet Grafana public.
Our happy-path blocks now publish ~0.2s into the slot, down from ~1s. That's within the first slot interval.
The change: a block-building optimization from a recent PQ interop call. We prepare blocks one interval earlier, doubling our budget from 0.8s to 1.6s.
Last week we flagged an error during internal interop runs: validators started generating useless votes when a fork advanced finalization without becoming canonical.
New blog post out: ethlambda, devnet 5 and beyond.
Each devnet introduced a new piece of the protocol's complexity. Devnet 5 is now live and in interop testing, and it's a good time to look back at how far we've come and what's on the roadmap ahead.