Lido on Ethereum: SSV Network Testing V2
During April through July, Lido DAO contributors collaborated with the SSV Network team and multiple Node Operators for a second round of testing of SSV Network-based Distributed Validators through the Lido Node Operator registry on Goerli.
The trial included 46 non-Lido Node Operators consisting of solo stakers, community stakers, and other professional organizations in addition to 10 existing Node Operators using the Lido on Ethereum protocol. This included operators running their SSV Operator through Dappnode, self-hosted servers, via colocation in datacenters, and public cloud.
This trial was an expansion of the first SSV Network integration on Goerli that featured eight existing Lido on Ethereum Node Operators. The goals of this round of testing were to trial operating SSV distributed validators through the Lido Node Operator registry with community stakers, and to form a foundation of what an initial limited mainnet application of SSV based DVT may look like.
Cluster Setup
For this trial, participants were separated into 10 clusters of participants, all of which are visible on operators.testnet.fi/. This included four clusters with a 3/4 threshold configuration, two with a 5/7, three with a 7/10, and the first 9/13 threshold cluster to be tested through the Lido registry.
Each cluster had a “cluster coordinator”, a member of the cluster responsible for setting up validator keys to split across the SSV key splitter tool and a corresponding multi-sig wallet that served as the entry to the Lido Node Operator register.
Of the 10 clusters, 3 clusters were led by existing Node Operators, and the remaining 7 were led by community stakers.
The clusters participating in the trial included: Cebu, Four Part Trilogy, Group 3 Avengers, AWD, BlueCrisps, Clusterix, The Magnificent Seven, Group 8, StakeVanguard, Group 10.
Two of the 3/4 threshold clusters were organized to trial an early implementation of a RockX-developed distributed key generation (DKG) ceremony mechanism. Although initially not successful and validators for these clusters were not activated, the testing process surfaced numerous bug fixes and process improvements, and it is estimated that a next round of testing would include successful DKG setups.
The remaining eight clusters (shown below) each activated and ran 5 validators.
Distributed Validator Setup
The cluster setup process started with each individual cluster member setting up their SSV Operator following the SSV operator instructions. Each participant created a brand new SSV operator specific to the trial.
Once done, each cluster set up a multi-sig which became the cluster’s entry in the Lido Node Operator Registry on Goerli. When each participant confirmed their operator was active and online, the cluster coordinator for the 8 non-DKG clusters was responsible for generating validator keys, submitting them to the Lido registry, and splitting them across their respective cluster members. At this point validator duties were then handled by each cluster’s distributed SSV operators.
Each of the 8 clusters successfully proposed blocks, demonstrated strong attestation performance, and excluding an initial performance issue related to peering, strong uptime.
Upon completion of the trial, cluster coordinators were responsible for exiting each validator which proved to be a seamless process.
Performance Metrics
While the setup process of the clusters and operators was mostly smooth, two issues were present at the start of the trial. 1. Cluster performance for those “greater than 7 operator” clusters was initially poor due to low peer connectivity, resulting in some validators failing to properly attest. The SSV team rolled out a new version (v0.5.4) that solved this issue and validators began properly attesting once all participants had updated their SSV operator software.
The second issue was an issue with SSV Validator registration, prohibiting two of the clusters from splitting their keys and registering to the SSV network. An update was rolled out to the SSV registration webapp that solved this issue soon after discovery.
For the 60 day sample period that validators were deposited to, the 8 clusters with active validators showed strong performance in-line to slightly above that of other validators run on Goerli per Rated Labs data (note that average validator performance on Goerli is much lower than mainnet), especially when considering only 8 active clusters and a connectivity issue that was subsequently fixed.
Analysis on validator performance from the 60 day period prior to the last validators exited (following the initial 60 day window above) shows clear improvement in performance without the skew of the initial setup issues.
For example, Average Attestation Rate* improved from 97.44% during the 60 day window to 99.04%. Once cluster setups were completed and operational, the validators maintained strong performance throughout the remainder of the trial.
Plans for Next Round of Testing
The second round of testing SSV based distributed validators was successful, clearly demonstrating the ability to combine solo stakers, community stakers, and professional operators across clusters through the Lido Node Operator Registry on Goerli in a performant manner.
In the next round of testing, a more mature DKG mechanism and blinded block support (to enable MEV-boost) are planned for testing.
As mentioned in the follow up post of the Obol Network trials, the Staking Router will potentially allow for new methods of participating in the Lido Node Operator Registry on mainnet in the coming months.
Potential methods (subject to a DAO vote) could include limited trials of simple distributed validator configurations on mainnet, more complex modules with native DVT-protocol mechanism integrations, and migration of stake from the single-operator model currently utilized in the Curated Operator Set.
The next round of testing is tentatively scheduled for early Q4 with both SSV Network and Obol Network.
If you are interested in participating in future Lido on Ethereum DVT testnets, please use this form to express your interest. In the coming months next steps will be announced.
Lido x SSV Testnet Participants
Group 1: RockX, DSRV, HashQuark, Cosmostation
Group 2: Cryptomanufaktur, Simply Staking, Eridian, H2O Nodes
Group 3: Kukis Global, Allnodes, Ebunker, Cnupy
Group 4: Anonstake, Astro-Stakers, Wiggyhop, ShardLabs
Group 5: RockLogic, P2P, Infstones, DSRV, Simply Staking, CVJointOps, Ellipfra
Group 6: Infstones, HashQuark, Cryptomanufaktur, Allnodes, Foundry Digital, SenseiNode, Cabinet42
Group 7: RockX, Kukis Global, P2P, Cryptomanufaktur, Infstones, Eridan, A41, Gateway.fm, ParaFi Capital, Tokenomist
Group 8: RockX, DSRV, HashQuark, Kukis Global, P2P, Hellman Research, Lydia Labs, OKX Pool, Posthuman, Chainbase
Group 9: 01node, Chainode, Spacesider, Spire Blockchain, Inc., Swiss Staking, DragonStake, Cypher Core, Anvil Finance, Lanski (Dappnode), Neuler FZCO
Group 10: Deutsche Telekom, Avaunt Staking, Luganodes, RockawayX Infrastructure, Metanull, Orion, Piconbello OU, Spectrum Staking, Validation Cloud, StakeWithUs Pte Ltd., Swifstaking, Stakeall Finance, Infinite Lux Staking Service