header-langage
简体中文
繁體中文
English
Scan to Download the APP
Partnered With:

Analyze Rollup fast finality and its application scenarios

Read this article in 5 Minutes
Some applications may require faster finality to achieve their functionality or performance.
原文标题:《 一文分析 Rollup 快速最终性及其适用的应用场景 》
Original article by Alex Beckett
原文编译:东寻,DeFi 之道


One of Celestia's main goals is to extend rollups by providing maximum data throughput. When scaling data throughput, the requirements of lightweight nodes can become bloated with faster block-out times. Because there is a trade-off between block time and light node overhead, a suitable medium needs to be found. For Celestia, the search for such a suitable medium resulted in the current 15-second block time.


As a result of a slightly slower chunk out time, rollups inherits its Finality. But some applications may require faster finality to achieve their functionality or performance. Rollups can only be as slow as Celestia's block time, right? Not really.


Single Sequencer


First, there are two concepts of Finality in rollup transactions.


Soft Finality: A collator or group of collators confirms that a user's transactions will be included ina block of Celestia.


Firm finality: Celestia confirms that a user's transaction is included ina final block.


The guarantee Soft Finality can offer users is that the collator promises to include your transaction ina block on Celestia. If they don't include users' transactions, then they will be penalized in some way.


rollups with a single collator now easily provide Soft Confirmations -- they are the only block producers. However, without an accountability mechanism to punish the sorter for dishonest behavior, Soft validation provides only a weak form of Soft Finality. Since few rollups have staked sequencer, they are not penalized if they soft-confirm a trade and do not include it in the next block.


However, a single collator is only an intermediate step. Soft validation for faster finality makes more sense (and is more interesting) in the context of decentralized rollup.


Decentralized sorter


There is a wide scope for designing the decentralization of collators. Now, I'll look at the soft validation of two decentralized collator mechanisms.


PoS leader election mechanism


For a quick background, as the name suggests, the way the PoS leader election mechanism works is that each node pledges a number of tokens to participate in the collator set. The sequencer agrees on the leader schedule in advance. When it's the turn of the collator to produce blocks, it builds its blocks on a schedule and submits them to Celestia. Once a new block is built, there is no BFT consensus among verifiers. We implemented something similar in the first release of the decentralized collator in Rollmint.


In this case, the collator can provide a soft acknowledgement to the user once the user's transaction builds a block. Soft confirmation is like the promise that they're going to include it in the Celestia block. Since the sequencer has now been pledged, they can be held liable if they provide soft confirmation and don't end up including the user's transaction in the next block.


One way soft validation works is that the collator builds a block and propagates it over the rollup's p2p network. At some point, the collator also publishes the block to Celestia. Nodes can verify that there is a mismatch between the blocks they verify on the p2p network and the blocks that are published to Celestia. The mismatch proves that the collator signed two blocks, so they can be penalized by cutting.


Currently, applications can get financial assurance behind their soft validation, allowing them to run with Fast Finality.


BFT consensus mechanism


Another mechanism for decentralized sorters is to use an existing Byzantine fault-tolerant consensus mechanism (BFT consensus), such as Tendermint. If Tendermint were used to provide Firm finality to rollup, its security would be no different from that of L1, which relies on honest majority consensus. rollup should use a modified version of Tendermint so that it is only responsible for soft validation. By doing so, rollup still relies on fraud/proof of validity and Celestia for security.


In this case, one collator builds a block and submits it to the rest of the collators. The collator votes the block in multiple rounds. If the user receives a soft confirmation of his transaction and it is not included in the Celestia block, then the vote is evidence of a double signature. Since they are also pledged, the sequencer may be cut.


The tradeoff between using PoS leader elections and BFT consensus mainly involves the overhead of the collator versus the execution time. The PoS leader election mechanism is more streamlined than the BFT consensus and has less overhead on the sorter. But decentralized sorters can be implemented faster through existing consensus mechanisms than customized leader selection mechanisms.


Which applications might need fast finality?


There may be several types of applications that require fast finality. Two come to mind:


1.DEX is the most obvious, because finality has an impact on the entire transaction dynamic. Faster finality allows trades to be cleared faster, allowing searchers to arbitrage more frequently, leading to better strike prices and more accurate asset prices for users.


2. The loan agreement may also want to have faster finality so that the liquidation can be processed more quickly. Faster liquidation reduces the time for asset prices to move, which means that lending agreements are less likely to produce bad debts. The longer it takes to complete the liquidation, the more likely it is that asset prices will fall, potentially rendering the agreement bad.


The final sex difference between 2 and 15 seconds probably won't affect most applications. Only a small percentage of applications may require fast finality, and most of them should fit the fast finality concept rollup can provide.


Original link


Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia

举报 Correction/Report
Choose Library
Add Library
Cancel
Finish
Add Library
Visible to myself only
Public
Save
Correction/Report
Submit