MAINNET:
Loading...
TESTNET:
Loading...
/
onflow.org
Flow Playground

Full Observer Node

Running your own full observer node


This is an alpha release and is known to contain bugs and is missing functionality. We are currently changing this codebase to address these issues. In the meantime please avoid building a dependency on this code until it's re-released to open beta.

A full observer node is an unstaked access node that follows the flow consensus and serves the Access API using local data. This allows anyone to setup an access node for local queries without the need for staking.

Running a node

The full observer node was built in collaboration with the Optakt team. The primary way to run a full observer node is with the Flow Data Provisioning Service (DPS).

Data Provisioning Service

The Flow DPS was built by the Optakt team and greatly streamlines use cases that require archival data. It provides a scalable and efficient way to access the history of the Flow execution state, both for current live sporks and for past sporks.

For more information on the Flow DPS, see their documentation.

The full observer node functionality is provided by the Flow DPS live indexer, which consumes live state updates from the Flow network by acting as a consensus follower.

Once fully synchronized, the Flow DPS live indexer serves the Flow DPS API for the active execution state in real time. The Flow DPS Access API can be run on top of the live indexer to enable use of the data for Access API clients.

Custom Implementations

Here is an example implementation using the consensus follower library: https://github.com/onflow/full-observer-node-example