Shared Sequencer

This comparison highlights the core trade-offs between the dominant models.

• Decentralization: Shared uses a validator set; Centralized is a single operator.
• Censorship Resistance: High in shared models; Low in centralized (operator can censor).
• Cross-Rollup Atomic Composability: Native in shared; Impossible in isolated centralized sequencers.
• Time-to-Finality: Can be slower in shared (consensus overhead); Often faster in centralized.

Building a production-grade shared sequencer involves solving several complex problems:

• Consensus for High-Throughput: Adapting BFT consensus (like Tendermint) to handle 10k+ TPS with low latency.
• Economic Security & Slashing: Designing staking, slashing, and reward mechanisms to deter malicious sequencing.
• Interoperability Standards: Creating universal APIs (RPC endpoints) for diverse rollup VMs to subscribe to the sequence.
• MEV Management: Implementing fair ordering or MEV redistribution (e.g., via MEV-boost analogues) at the sequencing layer.

A core security consideration is the decentralization of the sequencer set. A centralized or permissioned set of operators can censor transactions or extract MEV. Solutions include:

• Proof-of-Stake (PoS) validator sets where operators stake tokens.
• Permissionless sequencing via a decentralized network of nodes.
• Election mechanisms (e.g., DPoS) to rotate sequencer duties. The goal is to achieve liveness and censorship resistance comparable to the underlying L1.

The shared sequencer must provide strong data availability (DA) guarantees for the transaction batches it produces. If data is withheld, rollups cannot reconstruct their state, breaking safety. Key mechanisms include:

• Posting data directly to a robust DA layer like Ethereum or Celestia.
• Implementing fraud proofs or validity proofs that can be verified even if the sequencer is malicious.
• Ensuring timely settlement on the destination L1 to finalize state.

Economic security is enforced through cryptoeconomic incentives. Malicious behavior by a sequencer operator (e.g., proposing invalid state transitions) is disincentivized by slashing their staked collateral. This creates a bond that can be forfeited. The security of the system is directly tied to the value of this staked capital and the cost of attacking the network.

A shared sequencer has a privileged view of cross-rollup transaction flow, creating a powerful MEV (Maximal Extractable Value) extraction point. Without safeguards, this can lead to transaction reordering and front-running. Mitigations include:

• Commit-Reveal schemes to hide transaction content.
• Fair ordering protocols (e.g., first-come-first-serve in a mempool).
• MEV redistribution mechanisms back to the rollup communities.

The network must remain live, producing new blocks even if some sequencers fail or act maliciously. This is achieved through Byzantine Fault Tolerance (BFT) consensus protocols. Key metrics include:

• Finality time: How quickly transactions are irreversibly ordered.
• Halt tolerance: The number of faulty nodes (e.g., 1/3 or 1/2) the network can withstand while remaining operational.

Using a shared sequencer shifts trust assumptions from a single, often centralized, rollup operator to the decentralized sequencer network and its consensus protocol. Users must trust that:

1. The sequencer network is honestly decentralized.
2. Its consensus is secure.
3. The DA layer is available. This contrasts with native sequencing, where trust is placed solely in the rollup's own operator, or based sequencing, where trust is placed in the underlying L1 (e.g., Ethereum).

A Sequencer is a node responsible for ordering transactions before they are submitted to a base layer (L1). It is the core component that a Shared Sequencer network decentralizes. Key functions include:

• Transaction Batching: Grouping multiple user transactions.
• Ordering: Determining the final, canonical order of transactions within a batch.
• Execution: Processing transactions to generate state updates.
• Data Submission: Posting transaction data and proofs to the settlement layer.

Proposer-Builder Separation (PBS) is a design pattern, pioneered in Ethereum consensus, that decouples the role of block proposal from block construction. This is a key architectural inspiration for Shared Sequencers.

• Builder Role: Analogous to a sequencer, it assembles transactions and creates an execution payload for maximum value (e.g., MEV).
• Proposer Role: Selects the most valuable payload to propose for consensus.
• Shared Sequencer Parallel: In PBS, builders compete; in Shared Sequencing, sequencers may cooperate or compete to provide ordering services to multiple rollups.

Cross-Rollup Atomic Composability refers to the ability to execute a transaction that depends on the atomic, all-or-nothing success of operations across multiple distinct rollups. A primary value proposition of Shared Sequencers is enabling this.

• The Problem: Without a shared ordering layer, atomic cross-rollup transactions are impossible, as one chain cannot guarantee another's state.
• The Solution: A Shared Sequencer, seeing all transactions for all connected rollups, can order a multi-rollup transaction bundle atomically, ensuring all parts succeed or fail together.

Maximal Extractable Value (MEV) is profit that can be extracted from block production via transaction inclusion, exclusion, and ordering. Shared Sequencers centralize and can mitigate MEV concerns.

• MEV Centralization Risk: A single-rollup sequencer can capture all MEV, creating centralization pressure.
• Shared Sequencing Approach: Can implement fair ordering protocols (e.g., first-come-first-serve timestamps) to reduce predatory MEV.
• MEV Redistribution: Some designs allow MEV captured by the shared network to be redistributed back to the constituent rollups or their users.

A Data Availability (DA) Layer is a blockchain that guarantees the publication and storage of transaction data so anyone can reconstruct state. It is distinct from but complementary to a Shared Sequencer.

• Sequencer Role: Orders and executes transactions, producing state diffs and data.
• DA Layer Role: Securely stores the underlying transaction data with high availability.
• Interaction: A Shared Sequencer typically submits its ordered transaction batches to a DA layer (like Celestia, EigenDA, or Ethereum) before or alongside settlement proofs.

The Settlement Layer is the base blockchain (typically an L1 like Ethereum) where rollups finalize their state and resolve disputes. The Shared Sequencer interacts with this layer.

• Finality Anchor: The settlement layer provides ultimate economic security and finality for rollup states.
• Sequencer Output: The Shared Sequencer submits state roots or validity proofs to the settlement layer for verification.
• Escrow & Slashing: User funds and sequencer bonds are often held on the settlement layer, with slashing enforced for malicious behavior.