Ethereum’s Holešky Testnet: A New Era for Development and Testing [UPDATED]

The Ethereum team is launching a new testnet named Holešky.
Our team regularly attended the Holešky launch coordination calls and kept an eye on the update, so in this article, we’ve been collecting everything already known.  

Ethereum Sandboxes Testnets (abbreviated from a “test network”) are used by developers to safely test protocol upgrades and smart contracts before deploying them on the main network. It provides a safe environment for testing in conditions similar to the live network.

Before, there were two public testnets in Ethereum:

1. Sepolia that allows developers to test contracts and applications.

2. Goerli that allows protocol developers to test network upgrades and stakers to test running validators.

But it wasn’t like this earlier. 

In autumn 2022, a Merge in Ethereum required the deprecation of multiple testnets. However, the short notice of only a few weeks caused frustration among application developers and infrastructure providers. 

Thus, to avoid this in the future, a team introduced a predictable Ethereum Testnet Lifecycle plan:

Open graph

According to the plan, Goerli entered the long-term support phase in Q1 2023. After the end-of-life date (planned for Q4 this year), it will no longer receive support from client teams. Therefore, the next testnet should be launched at the end of 2023 to prepare for the Goerli shutdown.

Successor Is Coming

As you already understand, Holešky is precisely the next testnet we’ve been expecting. It’s not just some average upgrade but the first long-standing, merged-from-Genesis, public Ethereum testnet.

Holešky replaced Goerli as a staking, infrastructure and protocol-developer testnet. At the same time, decentralized applications (DApps) testing, smart contracts, and other EVM (Ethereum Virtual Machine) functionality will be available for testing on the Sepolia testnet.

Designed as a replacement for Goerli, Holešky allows public validators to join. This new testnet can be especially useful for node operators and DVT (Distributed Validator Technology) projects to test publicly.

Naming Holywar

The Holešky network is named after Nádraží Holešovice, a train station located in Prague, Czech Republic. Similarly, Goerli was named after Goerlitzer Bahnhof station in Berlin, commonly referred to as “Goerli” among Berliners.

However, the name “Holešky” is neither a direct abbreviation of Holešovice nor its logical adoption as with Goerli. So, at first, it was “Holli,” which was perceived as inappropriate by the Czech Ethereum community, as Holešovice is not just a train stop but an entire district with a rich history dating back to the 12th century.  

After engaging in discussions with local folks, the Ethereum team finally chose the name “Holešky,” which is an abbreviated form of Holešovice and pays tribute to its Prague heritage.

Estimated Timeline

To replace Goerli by the end of this year, the Ethereum team proposed launching Holešky no later than September.

Open graph

Thus, the Genesis event is preliminary scheduled for September 15, 2023, at 14:00 GMT. A live event in Berlin will be announced once the Genesis date is finally confirmed.

September 15 is not a random pick. This day is the 1st anniversary of The Merge, which completed Ethereum’s transition to Proof-of-Stake consensus and reduced energy consumption by ~99.95%. Read more about this significant milestone here and in our FAQ.

Here is a rough timeline of the launch represented on the second coordination call:

Current status

Coordination call #1

So, a first breakout call took place on June 15 to kick off discussions on a final Genesis configuration and state. We’ve participated along with  Ethereum Foundation devs, client teams, node operators, L2 applications, and Ethereum enthusiasts. 

Participants went through the purpose of the testnet, the date the Genesis would start, how big and how long it should be, and who will run the keys.

Key takeaways from coordination call #1

Here’re the major points from the discussion; a full transcript of the call is available here.

• The Genesis version is Shanghai/Capella upgrade, which means that Holešky will be in the same state as the Ethereum mainnet from Genesis.
• Devs will employ some supply tricks, the same as we have in the Sepolia testnet, to facilitate inflation more conveniently.
• If Holešky is big enough, core devs can adequately test new network upgrades before they go live on the Ethereum mainnet, for example, EIP-4844 (a proposal that introduces a new transaction format for “blob-carrying transactions” which contains a large amount of data that cannot be accessed by EVM execution, but whose commitment can be accessed), P2P stack, and others.
• As soon as they need to keep it bigger than the mainnet, there should be at least 1-1.5 million validators to start off at the current growth rate. Finally, participants agreed that a Genesis validator set size should be close to 1.4-1.5 million validators. However, client teams should consider if there can be any issues with such a large validator set.
• Client teams should ideally run at least 66% of the keys. The rest would be node operators, DVT, and other ETH stakers. This requirement was implemented after the Shanghai/Capella upgrade on the Goerli testnet. The lack of chain finalization was observed due to non-upgraded nodes, which held a portion of the testnet nodes.
• Goerli testnet experienced finalization after offline operators upgraded, which took approximately 15 epochs or 1.5 hours. To avoid similar situations during the Holešky Genesis event, developers aim to ensure the majority of validators are upgraded and operational. This will create a stable testnet, working like the mainnet.
• Recommended validator set size:
  – 80-100K for client teams. Ethereum EL/CL teams are welcome to commit here;
  – There is no recommended size for node operators, layer-2s, ETH stakers and other enthusiasts, but it would be suggested to commit to at least 10k keys. Node operators and layer-2s can commit link, and a separate one for ether stakers and other enthusiasts. 
• It should be considered conducting 200k exits right after the Genesis to get a more interesting consensus state.
• The majority of validators should be of type-0 withdrawal credentials.
• Holešky won’t replace Goerli entirely when launched. It will be deprecated and slowly shut down, so the budget for Goerli nodes running can be redirected to Holešky.
• The estimated lifetime should be 3-5 years.

Coordination call #2

The second call took place on June 29 and showed that the team is moving quickly, overcoming the planned milestones and will probably launch the update even earlier than planned.

Key takeaways 

• The required commitment of 1.5 million validators is already reached and even overpassed with 1.7 million validators aboard. The team aims to close the signups soon. The client teams are actively working on reaching a count of approximately 800,000.

• Ethereum Foundation plans to allocate a Genesis ETH size of a few hundred million ETH. This allocation ensures that there will be enough ETH for faucets and validators without worrying about supply overshoot issues.

• The allocations have been set similarly to the Genesis allocation tracker eth-clients/holesky#1 but with modified values that are more reasonable.

• A rough launch timeline was changed slightly; we’ve updated it for you in the section above. Now the launch is planned to be live in mid-September instead of October 1.

All Core Developers Execution call #167

All Core Developers Execution calls are fundamentally different from coordination calls. The Execution Layer Meeting discusses changes and challenges to the Execution Layer on Ethereum, where everything happens. This call, where the Holesky update was discussed, took place on August 3, 2023.

Key takeaways

• Parithosh Jayanthi, DevOps at Ethereum Foundation, and his team experimented involving large validators, specifically nodes operating 5,000 validator keys.

• Lately, they set up a testnet with 2.1 million validators distributed across 420 nodes. They quickly reached nodes’ CPU limits, prompting developers to replace the hardware with more powerful machines.

• The ultimate goal for the Holesky testnet is to support 1.5 million validators. Jayanthi expressed uncertainty about whether achieving this goal is possible based on the results of their experiments. He requested support from developers on the call to help assess the state of their preliminary networks.

All core developers consensus call #115

The Consensus Layer Meeting aims to discuss changes and challenges to the Consensus Layer on Ethereum, respectively. This one, which raised the Holesky launch update issues, happened on August 10.

Key takeaways

• Ethereum developers conducted experiments before the final closed launch of the Holesky testnet, particularly with the validator set size. Parithosh Jayanthi stated that the testnet initially had 2.1 million active validators, which appeared to be too much. This configuration led to issues with finalization due to too many validator duties and late block arrivals, resulting in insufficient attestations.

• To resolve these issues, the testnet was reduced to 1.4 million validators, and the maximum number of validators per node was decreased to 3,300 from 5,000.

• After adjustments, the testnet experiment succeeded, with block proposals ranging from 80% to 90% and attestations from 82% to 84%.

• The Holesky testnet will launch with an active validator set size of 1.4 million.

Ethereum All Core Developers Execution Call #168

This one happened on August 17.

Key takeaways

• Developers confirmed that the starting validator set size for the Holesky testnet will be 1.4 million validators. Before the decision, there were several weeks of testing large validator set sizes, as described above.

• Core developers decided to proceed with a default 1.6 billion ETH supply since testing had already been configured to support a larger ETH supply at Genesis.

• The team approved that the Holesky testnet launch was scheduled for September 15, 2023, because it’s the first anniversary of Merge activation.

Ethereum All Core Developers Consensus Call #116

As a final point of discussion on Ethereum All Core Developers Consensus Call #116, Barnabas Busa, DevOps Engineer at Ethereum, ran through three outstanding issues for the launch of the Holesky testnet.

Key takeaways

• Developers agreed to adjust the epoch time for the Shanghai/Capella upgrade on Holesky from 10 to 256 in an effort to mirror the environment of the Ethereum mainnet more accurately.

• They settled to use a validator ejection balance of 28 instead of 16.

• To complete a checkpoint sync of the Holesky testnet, developers agreed to support a compressed version of Genesis state.

15 September: Unsuccessful launch attempt

When the day came, the Holesky network experienced an improper launch due to a misconfiguration in the ExtraData field of the Execution Layer Genesis file. Although some validators tried manual fixes, the network couldn’t be finalized.

Currently, a final decision is pending, but there’s a big chance that Ethereum DevOps will relaunch a network using new genesis files within two weeks. Also, they intend to improve validation procedures and documentation to prevent similar issues in the future.

28 September: Successful launch 🎉

Ethereum developers made the second attempt to launch Holesky on Thursday, September 28. It wasn’t smooth sailing from the beginning, as the testnet didn’t reach finalization during the initial epochs.

Nevertheless, by epoch 5 of the Ethereum testnet’s relaunch, we observed a 66.6% finalization rate, meaning the network’s successful completion. As more validators have become active, the percentage has been steadily increasing.

Open graph

In the subsequent epochs, most validator groups have come online and are actively validating. 

Currently, the participation rate exceeds 90%. View here for real-time monitoring: Holesky Live Monitor.

You can watch how the launch took place here: EthStaker Holešky Testnet Launch Call.

To sum up

Holešky is a notable milestone as the first long-standing, merged-from-Genesis, public Ethereum testnet. It will take over Goerli as the primary testnet for staking, infrastructure, and protocol development. Meanwhile, the Sepolia testnet will remain available for testing decentralized applications (DApps), smart contracts, and other Ethereum Virtual Machine (EVM) functionality.

Holešky primarily benefits node operators and Distributed Validator Technology (DVT) projects, providing a public testing environment.

Although initially scheduled for launch on September 15, 2023, a misconfiguration in the ExtraData field of the Execution Layer Genesis file caused a delay. In less than two weeks, on September 28, Holesky was launched successfully and is live now! A big shout-out to the Foundation, client teams, and operators who have committed significant resources to ensure the successful launch of this chain.

Stake your ETH on Everstake with our newly launched staking solution, allowing everyone to stake starting at just 0.1 ETH. Staking takes just three clicks while benefiting your crypto holdings growth for as long as you wish.