Ethereum nodes can now remove certain historical blockchain records to...
Who Preserves Blockchain History When Nodes Stop Keeping It?
TL;DR
- Ethereum nodes can now remove certain historical blockchain records to reduce storage requirements, while continuing to validate new transactions.
- Historical records remain important for financial applications, analytics, compliance, and security investigations.
- As networks become more efficient, reliable access to older data will depend increasingly on specialized archives, data providers, and decentralized storage systems.
A blockchain is often described as a permanent record of transactions. Every transfer, smart contract interaction, and recorded event becomes part of its history, creating a record that applications can use to verify what happened.
However, maintaining that history comes with significant storage costs.
As blockchain networks grow, requiring every node to store every historical transaction becomes increasingly expensive. Ethereum has already started addressing this problem by allowing nodes to remove certain older records without affecting their ability to process new transactions.
The decision makes running infrastructure more practical, but it also raises an important question for companies building on blockchain networks: who will preserve historical information when ordinary nodes no longer keep all of it?
Why Ethereum Started Removing Historical Data
In July 2025, the Ethereum Foundation announced support for partial history expiry across its execution clients.
The change allows operators to remove block bodies and transaction receipts from before Ethereum’s transition to proof of stake in September 2022.
Previously, nodes maintaining complete chain history needed to retain large amounts of historical information, even when that data was unnecessary for validating new transactions.
According to the Ethereum Foundation, partial history expiry can reduce storage requirements by approximately 300–500 GB.
This improvement matters because operating blockchain infrastructure requires substantial disk space, bandwidth, and computing resources.
Ethereum’s October 2026 documentation reports that a Geth node retaining complete chain history can occupy more than 1.8 TB of storage.
Reducing unnecessary storage can make it easier for developers and independent operators to maintain nodes.
However, partial history expiry has limits. The broader proposal to automatically remove older history on a rolling basis, known as EIP-4444, remains under discussion.
Why Historical Data Still Matters
Although older blockchain records are unnecessary for many everyday transactions, businesses often depend on them.
Consider a fintech company building a transaction monitoring platform. Its software may need to reconstruct account activity from several years ago, identify transfers between wallets, or investigate a suspicious transaction.
A blockchain analytics company might use historical data to understand how users interact with decentralized applications. Security researchers may examine older transactions to identify the sequence of events leading to an exploit.
Financial institutions also need reliable records for reporting, reconciliation, and compliance.
These activities require different kinds of information.
Transaction history shows what was submitted to the network. Transaction receipts provide information about execution results and events. Historical state records allow developers to investigate what an account or smart contract looked like at a particular block.
An ordinary full node does not necessarily provide quick access to every historical state. Specialized archive nodes retain historical state information specifically to support these requests.
Understanding these differences is essential when designing applications that depend on older blockchain data.
Where Will the History Be Stored?
Ethereum’s approach does not require historical information to disappear from the ecosystem.
Instead, responsibility for preserving it can be distributed among different infrastructure providers.
Archive nodes are one option. These systems retain additional historical state information, allowing developers to retrieve older account balances, contract storage values, and other past records more efficiently.
However, archive nodes are more demanding to operate than ordinary nodes.
Ethereum’s documentation estimates that archive storage requirements can range from approximately 2 TB to more than 12 TB, depending on the client and its configuration.
Another approach involves specialized data providers that maintain historical records and offer access through APIs.
These services can simplify development, but relying entirely on one provider introduces questions about availability, pricing, and long-term access.
Ethereum also supports community-distributed historical archives through formats such as Era files, which package historical data for storage and retrieval.
The important consideration is whether information remains available, verifiable, and accessible when applications need it.
The Role of the Portal Network
One initiative addressing historical data access is the Portal Network.
Portal is designed to distribute blockchain information across many lightweight participants instead of requiring each participant to store the entire history.
Different nodes can hold smaller portions of historical data. When an application needs a specific record, the network can locate participants storing that information.
This approach could help reduce dependence on centralized infrastructure providers while making historical data more accessible.
However, availability still depends on enough participants preserving and serving the necessary information.
A decentralized network cannot retrieve information that nobody has retained.
For companies using blockchain data, the practical question is how reliably their chosen infrastructure can recover specific records during normal operations and unexpected failures.
Rollup Data Creates Another Challenge
Historical data preservation also matters for Ethereum’s Layer 2 networks.
Rollups process transactions outside Ethereum’s main execution layer and publish data to Ethereum to support verification.
Since the Dencun upgrade in March 2024, rollups have been able to publish certain information through temporary data structures called blobs.
These blobs reduce data publication costs, but Ethereum does not require nodes to retain them indefinitely.
Under EIP-4844, the required blob retention period is approximately 18 days.
After that period, nodes may remove the data.
This does not automatically make older rollup transactions invalid or erase the rollup’s state. However, applications needing the original blob contents must obtain them from another source if ordinary Ethereum nodes no longer retain them.
Rollup operators, historical data services, and independent archival systems therefore play an important role in long-term data availability.
What Companies Building on Blockchain Should Consider
Historical data access should be part of infrastructure planning from the beginning.
A company building an exchange, financial application, or analytics platform needs to understand which information its systems will require months or years later.
For example, a payment provider may need complete transaction records for financial reconciliation. An analytics company might need historical smart contract states to calculate past balances accurately.
Engineering teams should identify which data must be retained, who maintains it, and how the information can be recovered if a provider becomes unavailable.
They should also consider whether records can be independently verified and whether their existing tools support the required historical queries.
These questions become particularly important as networks continue reducing the storage burden on ordinary nodes.
Conclusion
Reducing blockchain storage requirements is an important step toward making decentralized infrastructure more accessible and efficient.
Ethereum’s history expiry developments demonstrate that networks can maintain their ability to process new transactions without requiring every participant to store every historical record.
However, historical information remains valuable to financial institutions, application developers, researchers, and security teams.
The long-term challenge is ensuring that this information stays accessible and verifiable as responsibility for storing it becomes more distributed.
Who We Are
Veretin Recruitment specializes in technical recruitment across Web3, AI, fintech, and blockchain infrastructure.
We work with companies to understand their products, technical requirements, and long-term goals. Through research-driven sourcing and careful candidate evaluation, we help organizations build strong engineering teams.
References
- Ethereum Foundation, Partial History Expiry Announcement, July 2025
- Ethereum.org, Statelessness, State Expiry and History Expiry
- Ethereum.org, Archive Nodes
- Ethereum.org, The Portal Network
- Ethereum Improvement Proposal 4444: Bound Historical Data in Execution Clients
- Ethereum Improvement Proposal 4844: Shard Blob Transactions
Originally published on Medium