Ethereum has officially confirmed that the Glamsterdam upgrade will activate on the Sepolia testnet on October 6, 2026, at 13:53:36 UTC, with the activation scheduled for epoch 353024 and slot 11296768. According to a proposal merged through Ethereum's protocol-management process, this represents the first major public test of the planned upgrade. The October 6 activation places Sepolia at the front of Glamsterdam's expected deployment sequence, with the upgrade's mainnet deployment still targeted for Q4 2026 according to the official roadmap. The deployment marks the first time Glamsterdam's code will run on a public Ethereum test network, a critical checkpoint before any mainnet launch can begin. The public test is the next major milestone following progress across Glamsterdam development and testing, with developers confirming the Sepolia date during the September 17 All Core Developers Consensus call. Recent market developments show ETH moving back above $2,609 on the Glamsterdam news, with short sellers scrambling as traders react to the confirmed timeline.
The upgrade is built around three primary aims: parallelization, expanded capacity, and preventing database bloating. As reported by 99Bitcoins, the Enshrined Proposer-Builder Separation (ePBS, EIP-7732) moves block handoff into the protocol, widening the block-propagation window from approximately two seconds to about nine seconds. Block-Level Access Lists (BALs, EIP-7928) provide validators with upfront transaction maps, enabling parallel disk reads and faster node syncing. Additionally, State-creation gas repricing (EIP-8037) targets predictable database growth of roughly 120 GiB per year. The upgrade aims to dramatically expand network capacity per block, with a target gas ceiling approaching 200 million, a substantial jump from the 60 million achieved in earlier testing phases. This change would allow more transactions and activity to fit into each block while keeping verification manageable. Glamsterdam follows Ethereum's earlier scaling work after Fusaka, with this upgrade focused on changing block construction, transaction processing and database management before testing moves from Sepolia to Hoodi.
The Glamsterdam upgrade addresses maximal extractable value (MEV), which Ethereum defines as value captured from block production beyond standard block rewards and gas fees through transaction inclusion, exclusion, or reordering. As reported by CoinDesk, EIP-7732 changes Ethereum's block-building architecture and reduces trust assumptions, but MEV matters because transaction order carries real financial value that often appears in final swap prices rather than visible blockchain charges. Users can protect themselves through several practical steps: avoiding setting slippage much higher than necessary, using deeper liquidity pools that experience smaller price movements, and employing private routing to keep transactions away from public mempools. Flashbots' Protect service shields Ethereum transactions from frontrunning and uses MEV-Share to return backrunning value to users. Intent-based trading systems offer another approach where users specify desired results and competing parties seek to fill orders, reducing harmful MEV exposure. The upgrade introduces protocol-level builder payments and removes reliance on third-party middleware for builder-proposer exchanges, fundamentally changing how builders and validators interact.
Developers have identified a critical security vulnerability during Sepolia testing that could undermine the entire purpose of the Glamsterdam exercise. As reported by BigGo Finance and confirmed during the September 17 All Core Developers Consensus call, consensus developer Potuz outlined a concerning attack scenario where an attacker could create a thousand fake builder identities, rotate through them, offer very high bids, win block auctions, and then simply refuse to hand over the promised transactions. The concern centers on free Sepolia ETH and disposable builder identities that remove financial friction for repeated bids, with the attacker's goal being to freeze Sepolia traffic and prevent other development teams from relying on the test network. "I can just spin up a thousand builders, rotate them, offer very high bids, and not produce payloads," Potuz warned during the call, adding "Any teenager can do this." The concern is disruption, not theft of real ETH, as those accounts could repeatedly win auctions for the right to provide the next block, then refuse to release the expected payload, reducing Sepolia's value as a realistic public test. This vulnerability adds another layer of complexity to the upcoming testnet deployment.
The path to Sepolia has been anything but smooth, with developers agreeing on a September 29, 2026 deadline for Sepolia-compatible client releases, leaving just seven days before the October 6 activation. This compressed timeline follows a string of earlier testing problems, including Devnet-8 experiencing finality issues during its Gloas activation and Devnet-9 hitting unplanned non-finality before teams eventually reached Devnet-11. The Hoodi testnet, which has an open validator set, is expected to follow with a tentative date of October 27, 2026, though its timing depends on how Sepolia performs first. The client teams have until September 29 to issue Sepolia-ready software, leaving seven days before the fork instead of the usual two-week window for security reviews, testing and bug fixes. Sepolia was chosen for the first public run partly because its validator set is controlled by a smaller group of teams, making it easier to coordinate a fix if something goes wrong, while Titan and Ultrasound have not yet completed Glamsterdam fork transitions.
