Author: 1912212.eth, Foresight News
As Ethereum’s Layer 2 networks mature and gas fees significantly decrease, these networks are showing a trend of rapid growth. However, accompanying issues such as fluctuating costs and transaction speeds that fall short of expectations have surfaced. Against this backdrop, the importance of transaction sequencing is becoming increasingly apparent as a key factor in resolving transaction bottlenecks and optimizing user experience.
In April this year, Astria, a blockchain module specializing in shared sequencers, completed a $5.5 million seed round with Maven 11 leading, alongside 1kx, Delphi Ventures, and Robot Ventures as participants. Just three months later, Astria secured another $12.5 million in funding, led by dba and Placeholder VC, with contributions from Hasu among others.
What is Astria?
Astria is developing a decentralized shared sequencer network designed to provide fast finality, censorship resistance, composability, and decentralization for Rollups.
Currently, using a centralized sequencer on L2 is more convenient, cheaper, and easier for users, hence most mainstream L2s manage their own teams. While L2 users can submit transactions directly to L1 to bypass the sequencer, they must pay L1 transaction gas fees and may experience longer transaction finalization times.
Sequencers control transaction sequencing and theoretically have the authority to exclude user transactions. Sequencers can also extract MEV from transaction groups. Centralization risks increase if there is only one sequencer. Therefore, a decentralized shared sequencer remains significant.
How Astria Operates
Astria’s decentralized sequencer comprises multiple sequencer nodes to sequence Rollup transactions. In Astria’s operational model, users submit transactions to Rollups, which automatically enter each Rollup node’s memory pool. Combiners collect these transactions and send them to the sequencer. Finally, the sequencer aggregates transactions into a shared block and sends pre-confirmations to users.
Current sequencers are specific to particular Rollups. Astria, however, processes blocks for multiple Rollups in batches. Data compression reduces costs when publishing data to L1. A decentralized shared sequencer network incentivizes participants from various Rollup ecosystems to act as validators on the network.
Astria Stack
Astria consists of five main components: Combiners, Sequencing Layer, Relayers, DA (Data Availability), and Scheduler.
Combiners
While technically adept individuals may directly use the sequencing layer for better transaction sequencing, it increases complexity for most ordinary users. Direct interaction with the sequencing layer requires users to hold sequencer tokens and maintain sequencer wallets, both negatively impacting user experience. Astria offers combiners to abstract this complexity—a gas station equivalent that manages transaction sequencing costs and provides unordered guarantees by bundling transactions as received.
Sequencing Layer
Astria’s sequencing layer uses CometBFT as its consensus algorithm. Chains supporting CometBFT can engage in Inter-Blockchain Communication (IBC), enabling cross-chain functionality.
A unique aspect of Astria’s sequencer is its deferred execution (delayed sequencing), where included transactions are assigned to another execution engine, Rollup. Sequencer nodes can opt to act as validators, actively participating in new block production and finality.
Relayers
Relayers retrieve validated blocks from sequencers and transmit them to the DA layer. Since sequencer blocks precede DA in terms of timing, relayers batch and compress ordered data from multiple sequencer blocks before submission to DA. Scheduler can fetch individual sequencer blocks for DA prior to submission, facilitating rapid finality improvements for enhanced user experience as a soft commit in the execution layer.
DA
Astria employs Celestia as its data availability layer, serving as the ultimate destination for sorted data from the sequencer network. Once written to Celestia, transaction order is considered final. Upon new Rollup node activation, all data is extracted from Celestia.
Scheduler
Scheduler functions as the consensus implementation for Rollup full nodes, akin to operation nodes in OP Stack. It bridges sequencers and the DA layer to the Rollup execution layer, extracting relevant Rollup data from each sequencer block, validating batched Rollup data, and transforming them into transaction lists for execution engines.
