Releasing Charon 1.0

Today, we’re releasing Charon 1.0, the first long-term support (LTS) version of our DV middleware client. Our development team spent the last three years building Charon, Ethereum’s first distributed validator (DV) middleware client. Today, after years of collaborative development, rigorous audits, thorough testing, and solid Mainnet performance, we have the confidence in Charon’s maturity to begin this next chapter of Obol’s journey. This first LTS version of Charon aims to remain compatible with future releases and serve as a foundation for large-scale deployments of distributed validators across the Ethereum staking ecosystem.

The Journey to Charon 1.0

In line with our design philosophy, we ensured that Charon’s debut on Ethereum’s testnets and Mainnet promoted not only decentralisation and resiliency, but also performance. A methodical, collaborative approach put Charon through two major testnet efforts, Athena and Bia, and two Mainnet phases, a closed Alpha and an Open Beta. In total, over 49,000 Obol DVs were run on testnet, and on Mainnet, DVs have achieved a cumulative 11,000 validator-days of runtime so far.  

Security & Audits

Charon, Obol Splits, and our DV Launchpad have completed multiple security assessments and audits to ensure the security of our architecture and code. The full list of audits is kept up to date in our docs, and is written here as of June, 2024:

To prepare for Mainnet deployments at scale, our core team has also published best practices guidance and a threat model documentation.

Secure, Resilient, Composable 

If Ethereum is to be the world computer, its consensus must be credibly-neutral and trust-minimised. This was the premise for our design and development of Charon: to prioritise security, resiliency, and composability. Optimising for these qualities led us to develop Charon as a DV middleware, the only one in the market today. By doing so, we’re confident that Charon provides the best security and performance for staking protocols and node operators, while maximising the decentralisation and client diversity of validators on the Ethereum network.

Some qualities of Charon include:

  • Distributed key generation (DKG): Charon’s DKG ceremony creates partial validator key shares for each of the nodes within a DV cluster, without ever creating a full private key at any point in time.
  • No runtime access to validator private key shares: Charon’s job is to reach consensus on what to present to the validator client to sign, and then to aggregate the returned signatures. Charon itself does not have runtime access to the validator private key shares nor the full private key. This minimises the risk of private key exfiltration through a vulnerability in Charon. 
  • Charon is compatible with any combination of existing clients: Charon is a DV middleware client that sits between the consensus and validator clients, allowing each node to use any combination of remote signer, validator, consensus, and execution clients. This builds client diversity into each validator, as each node in a DV cluster can use its own unique client combination.

If you’re interested in learning more, we’ve written extensively about Charon’s attributes and more in our blog. 

Charon’s Mainnet Journey

While we’ve shared updates over time about the development, testing, and initial adoption process of Charon, we never brought it all together in a single place. With today’s release of Charon 1.0, we’d like to present a timeline of how we got here:

Charon was gradually opened to the public starting in 2023, and in 2024 was quickly integrated by Lido and EtherFi to diversify and strengthen their node operator sets, committing $1.5b worth of stake to Obol DVs. But a lot happened in between:

  • First Mainnet DV: The first DV went live on Mainnet in December 2022, run by home nodes of Obol team members in Ireland, Estonia, and Canada.
  • Mainnet Alpha:  As part of our “Mainnet Alpha” phase, we invited the community to deploy the first set of DVs on Mainnet. From April - September 2023, we saw participation from 45 organisations and 27 solo operators, as described in our two blog posts. 
  • EtherFi Operation Solo Staker: An effort undertaken by EtherFi in collaboration with DappNode and Avado saw the participation of 40+ solo stakers from 35+ countries, running both bonded and unbonded DVs, as detailed in their blog post
  • Mainnet Beta: We officially entered our “Mainnet Beta” phase in December of 2023, enabling users of our DV Launchpad to configure clusters of 1 validator each on Ethereum Mainnet. Until now, over 1200 unique wallets have run DVs.  
  • EtherFi 1000 validators: EtherFi has deployed 1,000 validators on Obol DVs, constituting over $100m, strengthening the security and resiliency of their operator set.
  • Gnosis Chain: For the first time, DVs are running on a chain other than Ethereum. In this initial phase, 5,000 Mainnet validators are being deployed by the Gnosis core team with 8 professional node operators across two DV clusters. Obol will continue to work closely with the Gnosis team to introduce squad staking, increasing the decentralisation of the Gnosis Chain.
  • Lido SimpleDVT: Satisfied by the three waves of testing on testnet, (detailed below) Lido is bringing online 26 DV clusters as part of their SimpleDVT module. A DAO proposal in May raised the limit to 4% of Lido’s stake, constituting $1.4b of ETH as of the time of writing.
  • Squad staking: During our Mainnet Beta phase, we established squad staking, whereby community and home operators run validators together as a group of people and machines, lowering financial and technical barriers to entry. 

Charon’s Testnet Journey

From 2022 to 2024, over 49,000 DVs were run on testnet as part of multiple testing efforts, staking 1.57m testnet ETH, equivalent to $5.5b on Mainnet.  These testing efforts validated that DVs running Charon could achieve performance in line with traditional validators, even when run by a mix of community and professional operators using a huge diversity of client combinations, hardware setups, and geographical locations. 

  • “Athena” community testing: Over 1,100 operators in 40+ countries ran 200+ DVs on Görli testnet in Q3 of 2022. They achieved 99% proposer effectiveness and other metrics in line with the network average. Read more in our report and blog.
  • “Bia” community testing: 1,600+ DV clusters were successfully run by 9,000+ nodes in Q1 2023, with participation from 60+ professional organisations, as described in our blog
  • Internal 1000 validator test: A 1,000 validator 4-node cluster was run by the Obol team as a test on Görli from Dec 2022 - Feb 2023, described in the 2nd half of the Lido pilot testing report
  • MigaLabs performance testing: During March - May 2023, MigaLabs tested 4-node, 7-node, and 10-node clusters, running 1,000 validators each on Görli. They found that the difference between Obol DVs and traditional validators was under 1% for validator duties, as described in the report.
  • Lido testing wave 1: An initial pilot test by Lido with 11 operators. From Oct ’22 to January ’23 on Görli testnet, one DV was run on a 4-operator cluster, with a 2nd DV run on a 7-operator cluster, as described in the report
  • Lido testing wave 2: The first large-scale testing effort from Lido. From March - May 2023 on Görli testnet, 54 operators ran a total of 10 DV clusters with 5 validators per cluster, as described in the report.
  • Lido testing wave 3: Lido’s final preparation for SimpleDVT on Mainnet, mimicking their intended setup. Carried out from Nov ‘23 - Jan ‘24 on Holesky testnet, 214 Node Operators ran 32 clusters of 7 members and 100 validators each. Performance exceeded the requirements set by Lido, as described in the report.
  • Enterprise testing: In late 2023, our node operator partners Blockdaemon, DSRV, HashQuark, and InfStones ran testnet clusters of 3 to 7 nodes and 2,000 to 5,000 validators per cluster, for a grand total of 40,000 validators. High performance was achieved for clusters with reasonable latency between nodes. Enterprise testing in 2024 is ongoing with node operators from the Liquid Collective.
  • Techne Credential: An ongoing effort to award credentials to operators demonstrating good performance on testnet (and Mainnet). The credential provides on-chain attestation of experience running DVs with good uptime and effectiveness, giving operators credibility to approach staking protocols like Lido and EtherFi for an opportunity to run external stake on Mainnet.

What’s Next? 1.0 Releases of the DV Launchpad, Obol Splits, and Obol SDK

Charon is the first product within the Obol product suite to have its 1.0 release. Below is an overview of the Obol product suite, which in addition to Charon includes our Launchpad, SDK, and Obol Splits.

Following Charon, we’ll be releasing 1.0 versions of the DV Launchpad, Obol Splits, and Obol SDK.  If you have questions about integrating DVs in your infrastructure, or staking product, feel free to reach out to our business team.

Making DVs the Standard for Ethereum

The release of Charon 1.0 marks the transition to the next phase of DV adoption, with distributed validators evolving from an early technology to a standard primitive benefitting any staking participant. For staking protocols and node operators, running distributed validators improves resiliency, security, and decentralisation. For community operators or solo stakers, squad staking with DVs lowers the barriers to entry for validating on Ethereum and creates new opportunities to earn delegated ETH from staking protocols. 

Thanks to the creation of the Obol Collective and the “1% for Decentralization” retroactive funding program, anyone running Obol DVs contributes to ecosystem projects building on DVs and strengthening the Ethereum consensus layer. Join us on our journey. 🤝