Lab: free tools · size first · stock scanner · crypto scanner · journal

General

Archive Node vs Full Node: Ethereum Storage Modes Explained

A full node stores current blockchain state plus recent history but prunes old state to save disk space. An archive node stores the complete historical state at every block, which allows queries like "what was address X's balance at block 12,000,000?" Archive nodes require 10-20TB+ of storage for Ethereum but are essential for analytics, explorers, and DApps that need historical state lookups.

A full node verifies every block and keeps the current state, but prunes old historical state it no longer needs. An archive node keeps every historical state too, so it can answer questions like "what was this address's balance at block X" without replaying the chain. That makes archive nodes far larger and slower to sync. Most wallets and apps only need a full node; block explorers, analytics tools and some debugging workflows need an archive.

Practical Implications for Developers and Infrastructure Teams

When designing backend infrastructure for a blockchain-based application, the choice between data-complete storage and lightweight validation shapes cost, latency, and query capability. Teams building analytics dashboards, compliance tools, or historical auditing systems require deep storage access going back to genesis. Development teams focused on transaction broadcasting or wallet operations can generally operate with pruned, current-state storage without sacrificing functionality.

Cloud providers and managed blockchain-as-a-service platforms have reduced the operational burden considerably. Delegating heavy storage requirements to a managed provider allows small teams to access rich historical data without maintaining expensive on-premises hardware. Cost-benefit analysis should weigh query frequency, data freshness requirements, and long-term storage growth rates before committing to a particular infrastructure strategy. For comparisons across consensus mechanisms and infrastructure considerations, visit the DennTech tools page or browse related technical guides on the DennTech blog.

Sizes and sync times

A regular node keeps the state for only the most recent blocks, around the last 128, and can rebuild older states by replaying transactions when asked. An archive node keeps all of them. Most clients need more than 12TB of fast SSD storage to run in archive mode, while Erigon can hold the same data in under 3TB. An archive sync replays every transaction since genesis, and on an ordinary home computer that first sync can take up to a month. Regular use, like sending transactions, deploying contracts or staking, never needs historical states at all.

If you only need an occasional historical query, renting access from a provider is usually cheaper than running an archive yourself. If your business depends on that data, run your own, and keep in mind that a provider can change prices or limits on you.