Ethereum Glamsterdam Delayed After Security Audit Reveals Nearly 5x Gas Cost Spike That Could Permanently Break Deployed Contracts

Ethereum developers called off the Glamsterdam upgrade's Hoodi testnet activation four days before its scheduled launch on October 26, citing a security audit showing that two gas-accounting proposals could raise the cost of a common storage operation from 22,100 to 110,020 gas — nearly five times the current rate. ChainSecurity, which published the analysis on October 13, identified 13 real-world contracts already deployed on Ethereum that could be permanently harmed by this change, with some facing the risk of having failed operations locked in as irreversible rather than retryable. No replacement activation date has been confirmed, and ETH traded at $2,333.58 at the time of writing — broadly stable despite the delay.

Key Facts

• Ethereum developers postponed the Glamsterdam upgrade on the Hoodi test network at an All Core Devs Execution (ACDE) meeting on October 22, 2026 — four days before the planned activation time of 17:42:48 UTC on October 26

• The delay was triggered by ChainSecurity's October 13 analysis flagging risks in EIP-8037 and EIP-8038, two gas-accounting proposals bundled into Glamsterdam

• EIP-8037 could raise the gas cost of creating a new Ethereum storage slot from 22,100 to 110,020 — a nearly fivefold increase — when a separate gas sub-limit is exhausted

• ChainSecurity identified 13 real-world contracts already deployed on Ethereum that are vulnerable to the changed gas mechanics, without naming them due to an ongoing responsible-disclosure process

• The postponement also cancelled a planned upgrade to Hoodi's gas limit to 200 million, which developers had scheduled alongside the Glamsterdam activation

• Glamsterdam successfully activated on the Sepolia testnet on October 6 and remains live there

• Ethereum researcher Christine D. Kim has advised node operators to maintain their current software versions until a new schedule is set

• No confirmed replacement date has been announced; December has been discussed but not finalised

• At publication time on October 25, ETH traded at $2,333.58 — up 1.87% over 24 hours but down 1.59% over seven days — with a market capitalisation of approximately $285 billion, per CoinPaprika

The Delay Decision: How Ethereum Developers Halted a Live Testnet Activation

The decision to halt Glamsterdam's Hoodi testnet activation was made during the All Core Devs Execution (ACDE) meeting on October 22, 2026. The ACDE is the recurring coordination call where Ethereum's execution-layer client teams — the developers responsible for the software nodes that process transactions and enforce the protocol's rules — align on the network's technical roadmap, discuss the status of pending upgrades, and make decisions about activation timelines.

The meeting is one of the primary governance mechanisms for Ethereum protocol changes. Unlike blockchains governed by a central company or foundation with unilateral authority over upgrades, Ethereum operates through rough consensus among client teams and the broader developer community, with the ACDE serving as a structured forum for that consensus to form. A decision to delay an upgrade made at ACDE represents genuine multi-client agreement — not a unilateral hold by any single team or the Ethereum Foundation.

Glamsterdam had already passed its first major testnet milestone. The upgrade activated successfully on Sepolia — Ethereum's more openly accessible public test network — on October 6, nearly three weeks before the Hoodi delay was announced. Hoodi was intended to serve as the second and final public rehearsal: a testnet environment that more closely simulates mainnet conditions, particularly around validator behaviour and staking mechanics, before the upgrade proceeds to the live Ethereum network.

ChainSecurity's analysis, published on October 13 — nine days after Sepolia's activation and nine days before the ACDE meeting — identified security concerns serious enough for the developer community to conclude that the four-day window remaining before Hoodi's scheduled October 26 activation was insufficient to resolve them safely.

The Technical Core: What EIP-8037 and EIP-8038 Are Trying to Do

To understand why the gas cost changes in Glamsterdam create the problems ChainSecurity identified, it helps to understand what EIP-8037 and EIP-8038 are designed to achieve — and why those goals involve repricing storage operations so significantly.

Ethereum's State Growth Problem

Every smart contract deployed on Ethereum, every token balance, and every piece of data stored by any application running on the network occupies space in Ethereum's "state" — the comprehensive database of all current information about the blockchain that every full node must maintain. Unlike transaction history (which can theoretically be pruned or stored elsewhere), the state must be held in its entirety by nodes that need to validate new blocks, because any incoming transaction might read or modify any part of it.

Ethereum's state has grown continuously since the network launched in 2015, and the pace of growth has accelerated as smart contract usage has expanded. A rapidly growing state creates real operational challenges for node operators: larger state means more storage requirements, slower state lookups, longer synchronisation times for new nodes, and greater hardware requirements to participate in network validation. These pressures are directly at odds with Ethereum's goal of maintaining a decentralised validator set accessible to ordinary participants with consumer-grade hardware.

EIP-8037 addresses this problem by repricing one of the primary mechanisms through which state grows: the creation of new storage slots. When a smart contract writes a value to a previously unused storage location — creating a new entry in the state database rather than modifying an existing one — it creates a new storage slot. Under the current rules, this costs 22,100 gas. Under EIP-8037, creating a new storage slot could cost 110,020 gas when a separate sub-limit for state-creating operations has been exhausted, making new slot creation approximately five times more expensive in that scenario.

The dramatic price increase is a deliberate throttling mechanism. By making state growth significantly more expensive in scenarios where many state-creating operations occur in a single block, EIP-8037 limits how rapidly Ethereum's state database can expand even as the network raises its overall transaction throughput — a goal central to Ethereum's scaling roadmap. EIP-8038 adjusts the related costs for accessing storage slots that already exist, completing the repricing of the two primary storage interaction operations.

These are legitimate and important engineering objectives. The state growth problem is real, and the repricing approach is a recognised mechanism for managing it. The issue ChainSecurity identified is not that the goals are wrong — it is that achieving them through gas cost changes creates unexpected and potentially dangerous interactions with contracts that were designed and deployed under the previous cost structure.

The Vulnerability: Why 110,020 Gas Breaks Existing Contracts

The core of ChainSecurity's concern is the interaction between EIP-8037's repricing and a class of smart contract design pattern that relies on gas cost assumptions for its safety guarantees.

Gas Estimation and the Margin of Safety Problem

When a smart contract developer writes code that calls an external contract or performs a storage operation, they must account for how much gas that operation will consume. This is not purely a cost question — it is a safety question. Many contracts implement logic that deliberately limits how much gas they forward to external calls, or that checks whether sufficient gas remains before proceeding with a critical operation. These patterns are fundamental defensive tools against several classes of attack.

Under EIP-8037's new gas accounting, a storage slot creation that previously cost 22,100 gas now potentially costs 110,020 gas. Any contract that:

• Allocates a fixed gas buffer assuming the lower cost

• Performs a gas availability check calibrated to the old price

• Forwards a limited amount of gas to an external call that then attempts to create new storage

...may find that what was a sufficient gas allocation under the old rules is now catastrophically insufficient under the new ones. The contract doesn't just slow down or generate an error in most cases — because gas shortfalls during smart contract execution trigger irreversible state changes.

How Insufficient Gas Becomes Permanent Failure

ChainSecurity identified the specific failure mode: when the new gas rules cause an operation to exhaust its gas budget, the outcome is not simply "try again with more gas." Instead, the failed operation can be recorded permanently on-chain as a failure, with the contract's internal state updated to reflect that the operation was attempted and did not succeed.

For contracts that implement retry logic — mechanisms that allow a failed operation to be submitted again — this is catastrophic. If the initial failure is recorded as permanent, the retry path becomes inaccessible. A bridge transaction that should be retriable may become permanently stuck. A payment that failed due to gas exhaustion may be recorded as unclaimable. A governance vote or time-sensitive operation may be permanently lost rather than recoverable.

The attack vector ChainSecurity describes — insufficient gas griefing — exploits this property deliberately. A malicious caller can supply exactly enough gas to pass initial checks but not enough to complete a critical storage operation under the new pricing rules. The result is a controlled failure that locks the target contract into an unrecoverable state without spending significant funds: the attacker achieves permanent harm at minimal cost. Under the old gas pricing, contracts could protect themselves against this with gas checks and safety margins. Under EIP-8037's new pricing, those same checks and margins may be calibrated incorrectly, leaving previously secure contracts newly exposed.

The 13 Affected Contracts: Responsible Disclosure in Practice

ChainSecurity stated that it identified 13 real-world contracts already deployed on Ethereum that exhibit vulnerability to these gas-related failure modes. The firm did not name the affected applications in its public disclosure, explicitly citing the ongoing responsible-disclosure process.

Responsible disclosure is the standard security research practice by which a researcher who identifies a vulnerability notifies the affected parties privately before publishing the details publicly. The goal is to give developers time to prepare — patching or migrating contracts, alerting users, or implementing other mitigations — before a public announcement that might attract malicious actors seeking to exploit the vulnerability before it is fixed.

In the blockchain context, responsible disclosure is complicated by the immutability of deployed smart contract code. A contract already on Ethereum cannot simply be updated with a security patch the way a software application can. Developers of affected contracts may need to deploy new contract versions, redirect liquidity or user activity, or implement emergency pause mechanisms — all of which take time and require user communication. The responsible-disclosure window ChainSecurity is maintaining gives those teams the time they need before the full details of the vulnerability are known publicly.

The Ethereum Foundation has been assessing the broader scope of potentially affected contracts beyond ChainSecurity's initial 13 examples, attempting to determine how many deployed contracts across the mainnet may exhibit similar patterns.

Glamsterdam and the Cancelled Gas Limit Increase

The Hoodi postponement had a secondary consequence beyond simply delaying the upgrade: it also cancelled a planned increase in Hoodi's gas limit to 200 million.

Ethereum's gas limit is a network-wide parameter that determines how much computational work can be included in a single block. The mainnet currently operates with a gas limit of approximately 60 million. Raising Hoodi's limit to 200 million — more than three times the mainnet figure — was intended to stress-test Glamsterdam's parallel transaction processing features under conditions significantly more demanding than the mainnet currently imposes, simulating a possible future configuration after the upgrade is live and the gas limit has been raised on mainnet as well.

The cancellation of this gas limit increase means that particular test scenario will not be conducted on Hoodi during the current cycle. Once a new activation date is set and the gas mechanic safety concerns are resolved, the gas limit increase is likely to be rescheduled alongside the upgrade.

ETH Price Response: Market Shrugs Off the Delay

Despite the regulatory and technical significance of postponing a major protocol upgrade, the ETH market response was measured. At the time of writing on October 25, 2026, ETH traded at $2,333.58 — up 1.87% over the preceding 24 hours, according to CoinPaprika. The seven-day performance was slightly negative at -1.59%, and Ethereum's total market capitalisation stood at approximately $285 billion.

The muted market reaction likely reflects several factors. Ethereum protocol upgrades and testnet delays are understood by sophisticated market participants as a feature of rigorous development rather than a signal of distress — the network's multi-testnet process exists precisely to surface issues like the one ChainSecurity identified, and delay at the testnet stage is far preferable to discovering the vulnerability after mainnet deployment. The upgrade has already demonstrated it can activate successfully on Sepolia, suggesting the core functionality is sound and the issue is specific to the gas accounting changes rather than Glamsterdam's other features.

The broader market context — a period of general volatility across risk assets including cryptocurrencies — also likely dominated short-term ETH price movements, making any signal attributable specifically to the Glamsterdam delay difficult to isolate.

What Happens Next: No Confirmed Date, December Possible

The Ethereum Foundation has not announced a confirmed replacement activation date for Glamsterdam on the Hoodi testnet. Developers discussed the possibility of a December window during planning conversations, but those dates have not received formal confirmation from the ACDE or been announced publicly.

The timeline for setting a new activation date depends on how quickly several parallel workstreams can be completed:

Contract safety assessment: The Ethereum Foundation's effort to determine the full scope of mainnet contracts exposed to the EIP-8037 and EIP-8038 gas mechanics vulnerability needs to reach a conclusion sufficient to either confirm the risk is manageable or identify changes needed to the EIPs themselves.

Responsible disclosure completion: ChainSecurity's ongoing responsible-disclosure process with the developers of the 13 identified affected contracts needs sufficient time for those teams to assess their exposure and implement any necessary mitigations before the vulnerability details become fully public.

EIP modification if required: If the contract safety assessment concludes that the gas pricing changes in their current form cannot be made compatible with the security requirements of deployed contracts, the EIPs themselves may need to be revised — a process that requires drafting, community review, and re-implementation in client software, adding additional time before any testnet activation can be rescheduled.

Client software updates: All major Ethereum client teams need to release software that includes the revised Glamsterdam specification (with or without EIP modifications) before a new testnet activation can proceed.

Christine D. Kim's public notice to node operators — advising them to keep their current software versions until the network establishes a new schedule — reflects that all of these workstreams need to reach completion before operators will need to take action. Operators who had already installed Glamsterdam-enabled client versions in anticipation of the October 26 activation should follow the guidance from their respective client teams on whether to roll back to earlier versions or await updated releases that remove the Hoodi activation.

Implications for Ethereum's 2026 Mainnet Target

Glamsterdam's developers had been targeting a mainnet activation before the end of 2026 — an ambition that the Hoodi delay now places under significant uncertainty.

The standard Ethereum upgrade pipeline typically requires a successful Hoodi testnet run before proceeding to mainnet. If the December window for Hoodi proves achievable — and if a new Hoodi activation goes smoothly — the mainnet target could theoretically still be reached in very late 2026, though the margin would be slim. A further delay in resolving the gas mechanic issues, or complications identified during a future Hoodi test run, would push the mainnet activation into 2027.

A 2027 mainnet activation would in turn affect the timeline for Hegotá — Ethereum's subsequent planned upgrade, which is expected to introduce native privacy features at the protocol level and is currently projected for 2027. Ethereum's development culture prioritises safety over speed, and the pattern of each upgrade informing the next means that delays compound through the roadmap in both directions.

The security-first decision to halt Glamsterdam at the Hoodi stage — rather than proceed and address the gas mechanic concerns post-hoc — is consistent with that culture and is the correct call given ChainSecurity's findings. The cost is timeline uncertainty. The benefit is that 13 deployed mainnet contracts — and potentially more identified through the ongoing assessment — avoid being exposed to gas-related permanent failure modes on the live network.

Summary

Ethereum's Glamsterdam upgrade stalled at the Hoodi testnet stage on October 22 after a ChainSecurity audit revealed that EIP-8037 could raise the gas cost of creating a new Ethereum storage slot from 22,100 to 110,020 in certain conditions — a near-fivefold increase that can cause deployed contracts expecting the old cost to run out of gas and permanently record failed operations rather than keeping them retryable. ChainSecurity identified 13 real-world examples and is managing responsible disclosure with affected developers. The delay also cancelled a planned Hoodi gas limit increase to 200 million that was intended to stress-test Glamsterdam's parallel processing capabilities. Glamsterdam remains live on Sepolia, but no replacement date for Hoodi has been confirmed; December remains a possibility. ETH traded at $2,333.58 at publication time with a market cap near $285 billion, reflecting a market that views testnet-stage delays as a normal and even healthy feature of Ethereum's rigorous upgrade process rather than a signal of fundamental concern.