Ethereum Foundation and other protocol-level developers have decided to launch a "shadow testnet" after an error brought down an existing network while testing Ethereum's latest greatest upgrade. The hope is to prevent a major delay in this upcoming hard fork, Pectra, which was initially planned for rollout in early spring.

On Feb. 24, Ethereum's Holesky testnet experienced a significant disruption during the Pectra upgrade testing that has still not been resolved. This means that a lot of Pectra research and fine-tuning has been derailed until at least the end of March, when the Holesky is expected to finalize and start producing blocks again.

At least, that would be the case. On an All Core Developer’s call on Thursday, leading Ethereum researchers reached consensus on a quick fix solution — launching a "shadow fork" of Holesky —  to prevent too much disruption to Pectra's rollout. It’s still unclear exactly when Pectra will officially roll out, but the hope is to prevent too much disruption.

"We have some testnets to get into," Alex Stokes, co-leader of the Ethereum Foundation's research department, said to kick off the ACD Call 152. "Holesky is not going to finalize until the end of the month … and is unusable right now."

The Holesky failure wasn’t connected to an issue with Pectra itself but rather due to a configuration issue between key pieces of client software that help run Ethereum. While a hotfix was quickly found, a second testnet error on Sepolia on Wednesday and the fact that Holesky is still broken has raised concerns that this much-anticipated upgrade will be delayed.

This all comes amid a period of disillusionment among some Ethereum community members, who argue that leadership — especially at the EF — has sacrificed competitive network development to focus on arcane technical and social issues.

The Pectra upgrade contains a number of changes for Ethereum's execution and consensus layers that will make the network easier to use and more efficient. This includes upgrades to enable users to pay for gas in tokens other than ETH, introducing "account abstraction" that essentially turns wallets into programmable mini-accounts as well as increases to staking limits and blob count.

In an effort to find a practical solution to recent hurdles, DevOps researcher at the EF Parithosh Jayanthi noted that a parallel version of Holesky could be rolled out by next week. "We need a testing location for the next couple of weeks," he said, adding that this clone would essentially have all the major features of Holesky.

This is not the first time Ethereum devs have used a "shadow fork" to test an upcoming upgrade, though it may come under less auspicious circumstances. In 2022, for instance, the EF created a shadow fork of the Ethereum mainnet to thoroughly probe the upcoming switch to proof-of-stake, a widely ambitious technical upgrade compared at the time to "swapping out a rocket's engine mid-flight."

However, some technical issues may come up with a new network — to say nothing of the costs associated with the endeavor.

Jayanthi noted that the shadow fork would likely have "the same number of validators" as Holesky, which has about 10 times less than Ethereum’s mainnet, but it's unclear if this second network would use a new validator set or clone the original. This could cause issues for some applications that have tailored code to run on Holesky.

Moreover, it's less than clear whether certain "dependencies" like oracles or instances of major applications like Aave would be able to rollout along the same accelerated timeline.

"For us, it’s really critical that it works," Ivan Metrikin, a core contributor to Lido, the largest decentralized staking solution on Ethereum, said on the call. "Lido has a lot of depedendcies [and] that second network would have to be not worse in terms of features."

"The most important for us is that we have a similar setup as mainnet in terms of at least in client diversity and the types of validators being run," Metrikin said, saying that the absolute number of validators doesn’t matter but rather the quality. 

There is also an open question about whether the Ethereum Foundation would look to maintain the second network after it serves its purpose for testing Pectra as well as a related question of whether to depricate Holesky once it finalizes. Jayanthi noted the organization can likely “run 1,000s of nodes for a month” and also suggested “repurposing” the Morden testnet.

Metrikin noted there aren’t "any blockers" to testing on the shadow fork, especially if the team can use Holesky once it comes back, but they won’t be able to keep team members running tests on both networks "indefinitely."

Others raised similar concerns about the viability of testing a testnet. A developer by the screen name “Potuz,” for instance, noted that there is a “tradeoff in rushing without really diagnosing the bugs” that appear on Holesky. “You can investigate a lot of things on a shadow fork that is stable, but definitely not things that depend upon decentralization.” 

“A shadow fork would only be a test, it wouldn't be a testnet,” Potuz said, adding that a lot of this depends on how many resources the EF can commit to making it a truly decentralized operation. Stokes said the developers could revisit the deprecating issue once Holesky is finalized.

Radek, another common presence on ACD calls, noted that there is a risk of Holesky becoming “a second-class citizen” if people begin to “abandon” Holesky. This raises issues for applications like Lido, for instance, which has a “bunker mode” that cannot be truly tested without the many special configurations of Holesky.

In the end, most agreed on spinning up the shadow fork and many argued that having an additional testnet wouldn’t be the worst thing. Despite initial skepticism, for instance, Radek said it makes sense to keep both Holesky and its fork around, especially if they’re “not 100% sure” what they test on the fork is up to snuff "to say 'let's go to mainnet.'"

"Basically the shadow fork will become the testnet for Pectra, and Holesky will become the testnest for Fusaka," he suggested, hinting at Ethereum's next planned hard fork.

Disclaimer: The Block is an independent media outlet that delivers news, research, and data. As of November 2023, Foresight Ventures is a majority investor of The Block. Foresight Ventures invests in other companies in the crypto space. Crypto exchange Bitget is an anchor LP for Foresight Ventures. The Block continues to operate independently to deliver objective, impactful, and timely information about the crypto industry. Here are our current financial disclosures.

© 2025 The Block. All Rights Reserved. This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial, or other advice.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。