We set up archive nodes for projects that need full access to historical blockchain data. You run eth_call on block 5,000,000 and get missing trie node — a full node does not store historical states. A standard full node holds only the current state and the last few thousand blocks, but when you need the balance at block 15,000,000 or the state of a smart contract two years ago, the full node is helpless. An archive node stores the state trie for every block, ensuring complete access to historical data.
An archive node is not a luxury but a necessity for blockchain analytics, auditing, and dApps that work with historical data. However, the disk requirements are colossal:
| Network | Full Node | Archive Node |
|---|---|---|
| Ethereum mainnet | ~1 TB | ~18+ TB |
| BSC | ~1.5 TB | ~3+ TB |
| Polygon PoS | ~600 GB | ~8+ TB |
| Arbitrum One | ~300 GB | ~4+ TB |
| Optimism | ~200 GB | ~2+ TB |
Why does an archive node need 18+ TB of disk?
Each Ethereum block contains not only transactions but also a state trie snapshot. An archive node stores this snapshot for every block, unlike a full node that keeps only the latest states. This allows eth_getBalance on any historical block, but at the cost of enormous disk space. For comparison, a full sync of Ethereum mainnet with Erigon takes ~3 TB, while with Geth it reaches 18 TB. Erigon uses staged sync, which drastically reduces the required space.
When do you really need an archive node?
Before deploying an archive node, check alternatives: for analytics, The Graph subgraph or Dune Analytics already have indexed data; for one-off queries, paid archive RPCs (Alchemy, Infura). But if you have regular high-volume production queries, your own node pays off: savings on RPC can reach $2000/month at 10M historical requests. Additionally, your own node provides low latency and confidentiality — data does not leave your infrastructure.
How to set up an Ethereum archive node with Erigon?
Erigon is an alternative to Geth that stores archive data much more compactly (~3 TB for Ethereum). According to Erigon documentation, archive mode is enabled with the --prune='' flag.
Installation:
git clone https://github.com/ledgerwatch/erigon cd erigon make erigon Running in archive mode:
erigon \ --datadir=/data/erigon \ --chain=mainnet \ --prune='' \ --http \ --http.addr=0.0.0.0 \ --http.port=8545 \ --http.api=eth,erigon,web3,net,debug,trace,txpool \ --ws \ --torrent.download.rate=512mb \ --metrics --metrics.addr=0.0.0.0 --metrics.port=6060 The setup process includes several stages:
- Server preparation: CPU 16+ cores, RAM 64 GB, disk 4 TB NVMe SSD, network 1 Gbps.
- Install and run Erigon with
--prune=''. - Wait for synchronization (5–14 days on NVMe). Erigon syncs in stages: Headers → Block Bodies → Senders → Execution (longest) → Hash State → Intermediate Hashes → History Index → Log Index → Tx Lookup → Finish.
- Verify functionality by querying an old block.
- Configure monitoring and backups.
How to choose between Erigon and Reth?
| Parameter | Erigon | Reth |
|---|---|---|
| Disk size | ~3 TB | ~2.5 TB |
| Sync speed | 5-14 days | 3-7 days |
| Stability | High | Medium (less mature) |
| Language | Go | Rust |
Reth syncs faster but is currently less stable. The choice depends on priorities. If stability is critical, go with Erigon; if speed matters, consider Reth.
Monitoring and backups
After setup, verify archive mode by querying an old block:
curl -X POST http://localhost:8545 -H "Content-Type: application/json" -d '{"jsonrpc":"2.0","method":"eth_getBalance","params":["0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045", "0xF4240"],"id":1}' Erigon and Reth export Prometheus metrics. Key ones: erigon_stages_progress, chain_head_block, system_disk_free. Set up an alert for free space below 500 GB — an archive node grows actively.
Backups are critical. We recommend snapshots via LVM or rsync to a separate server. For example, after stopping the node: rsync -avz /data/erigon/ backup:/backup/erigon/. Incremental backups via rclone to S3 are cheaper than full copies.
What is included in turnkey archive node setup?
We offer:
- Hardware configuration selection tailored to your budget and load.
- Installation and setup of the node (Erigon, Reth, or L2).
- Monitoring and alerting configuration.
- Automatic backup setup.
- Load testing and optimization.
- Documentation and training for your team.
- Stable operation guarantee and support during failures.
Timeframe — from 3 to 10 days depending on the network and scope.
Our experience
Over 5 years in the market, 50+ configured nodes for clients. We know how to save costs without sacrificing reliability. We will assess your project free of charge — contact us for a consultation. Get a turnkey solution with documentation and support.