Snapshots and Network State

Overview

On the Hypergraph, snapshots serve as the core unit of state progression, replacing the traditional blockchain model of sequential blocks. Instead of processing transactions individually within a single chain, the network allows multiple L1 layers—such as DAG L1 and Currency L1—to create blocks independently in parallel. These blocks are then submitted to the L0 layer (Hypergraph/metagraph L0), where they are validated and aggregated into snapshots. Each snapshot finalizes a set of blocks, creating a cohesive and secure record of network activity.

This snapshot-based approach enables high throughput by allowing concurrent execution of L1 blocks while maintaining strong security guarantees through the Hypergraph’s global finalization process. Snapshots provide a scalable way to track and verify state changes across both the Hypergraph and metagraphs.

The Role of Snapshots on the Network

A snapshot in Constellation Network serves a function similar to blocks in other blockchains, but with key differences. Rather than containing individual transactions, a snapshot finalizes a collection of validated L1 blocks. These snapshots form a cryptographically linked chain that represents the history of the network’s state.

Each Global Snapshot records the Merkle root of all finalized transactions, along with references to previous snapshots to maintain continuity. It also includes any metagraph snapshots that have been submitted to the Hypergraph, ensuring that metagraph state is secured at the L0 level. Once finalized, a snapshot is immutable and verifiable, creating a reliable source of truth for all network participants.

The Snapshot Lifecycle

The process of snapshot creation follows a structured lifecycle, starting with L1 block formation, progressing through L0 validation and aggregation, and concluding with finalization and inclusion in the snapshot chain.

At the L1 level, multiple independent layers operate in parallel, each producing blocks that contain transactions relevant to their domain. For example, DAG L1 processes standard network transactions, while Currency L1 handles token-based operations. Validators in each L1 execute transactions, package them into blocks, and submit them to the L0 layer.

Once L1 blocks reach the L0 layer, they undergo validation to ensure correctness and consensus agreement. The Hypergraph’s L0 nodes verify the integrity of these blocks, confirm their inclusion criteria, and aggregate them into a snapshot. Each snapshot contains a finalized record of all validated L1 blocks within that period, along with a Merkle root summarizing the network state.

After validation, the snapshot is proposed to the network for consensus. Once finalized, it is added to the snapshot chain, permanently recording all included transactions and metagraph state updates. This process repeats continuously, allowing the network to progress efficiently without relying on a rigid sequential block structure.

Snapshot Triggers

Snapshot creation can be triggered in two ways: on-demand or at timed intervals.

On-demand snapshots are triggered whenever new L1 blocks are received by the L0 layer. As soon as enough data is available, the network initiates a round of consensus, ensuring that transactions are processed as quickly as possible. This mechanism allows for rapid state updates and fast transaction finality times.

Timed snapshots, on the other hand, are produced at regular intervals of approximately one minute, regardless of whether new blocks have been submitted. These timed snapshots serve an additional purpose beyond finalizing transactions—they increment the epochProgress value, providing a rough measure of time within the network. This concept of network time is crucial for operations that depend on periodic updates, such as validator reward distribution. Unlike on-demand snapshots, which are created as needed, timed snapshots ensure that certain protocol mechanisms execute at predictable intervals.

Metagraph Snapshot Lifecycle

Each metagraph follows a similar process to the Hypergraph for state progression. Metagraphs operate their own L1 layers, where transactions specific to that metagraph are processed and finalized into blocks. These blocks are validated by the metagraph’s L0 layer and aggregated into metagraph snapshots.

To ensure security and finality, metagraph snapshots are periodically submitted to the Hypergraph. Once included in a Global Snapshot, they become part of the immutable network history, securing metagraph state alongside Hypergraph state. This hierarchical model allows metagraphs to maintain autonomy while still benefiting from the Hypergraph’s global validation and consensus.