Archival Node Operation

Some blockchain queries require historical data that standard full nodes don’t retain. Archival nodes store the complete chain history, enabling queries against any historical block height.

What Archival Means

A standard full node prunes old state to save storage. An archival node retains everything — enabling queries like “What was this wallet’s balance at block 10,000?” or “What was the contract state 6 months ago?”

How PATH Detects Archival

PATH’s EVM QoS module automatically tests whether Supplier endpoints can serve historical data. It queries a known old block height and checks whether the endpoint returns valid data or an error.

• Archival endpoints are marked is_archival: true and can serve historical queries
• Non-archival endpoints are excluded from requests requiring historical data
• Detection expires periodically and is re-validated (see expires_at in the /ready?detailed=true response)

Check endpoint archival status:

curl "http://localhost:3069/ready/eth?detailed=true" | jq '.endpoints[].archival'

Which Chains Need Archival

Not all chains require archival nodes. It matters most for:

• EVM chains (Ethereum, Base, Arbitrum, etc.) — DeFi apps, analytics tools, and block explorers frequently query historical state
• High-value chains where historical data queries are common in production workloads

For lower-traffic or newer chains, standard full nodes are usually sufficient.

Storage Requirements

Archival nodes require significantly more storage than pruned nodes:

Running an Archival Node

1. Choose your chain — focus on chains where archival queries generate meaningful relay volume
2. Provision storage — NVMe SSD, sized for the chain’s full history plus growth headroom
3. Sync from genesis — or use a snapshot with full state history (check chain-specific resources)
4. Configure your backend — disable pruning in the node’s configuration
5. Register the service — include the chain’s service ID in your Supplier stake config
6. PATH auto-detects — once your node is serving historical queries, PATH marks it as archival

Compute Unit Considerations

Archival queries may consume more compute units than standard queries due to their complexity. The ComputeUnitsPerRelay for a service is the same regardless of whether the query is archival — but archival-capable nodes may receive more complex (and thus more valuable) relay sessions.

Adding Archival Detection for New Chains

If you’re extending PATH’s QoS to support archival detection for a new chain type, see the PATH repository’s qos/ directory. The EVM archival implementation (qos/evm/) is the reference example — it queries a known historical block and checks for valid responses.