Approved | Special Proposal to Metis Community: Migrate Data Availability to Ethereum Mainnet

Author: Metis

Date: 2023-09-27


Metis proposes to change the method it uses to process transactions from Layer 1 Ethereum – from an off-chain storage layer to convert to its original traditional Optimistic Rollup format.


In the spring of 2022, Metis partnered with a project called Memo Labs to create an off-chain storage layer. This storage layer has enabled Metis to avoid having to push all data to Layer 1 Ethereum; fewer data pushed to Ethereum has led to lower cost to transact on the Metis Layer 2 network.

Migrating to an Optimistic Rollup architecture will simplify the launch of a breakthrough venture: the Metis sequencer pool. Pushing all transactions to Ethereum streamlines the entire process, which would make it easier for Metis developers to launch this sequencer pool.

An important function of Optimistic Rollups is a concept known as data availability. Data availability ensures that the network’s transaction data are available to be reconstructed by validators and users at any time. This is typically done by batching transaction data to Layer 1.

Relying on a third party such as Memo Labs can have an adverse impact on data availability, as not all transaction data are published to Layer 1 – only the location of the data stored in Memo Labs. Migrating to Metis’ original Optimistic Rollup format would mark a heightened commitment to data availability, and to Ethereum’s ethos, while still enabling transaction fees that are in line with other Layer 2 networks, and much less expensive than on Layer 1 Ethereum.

Metis Sequencer Pool

Metis will soon become the first Layer 2 network to decentralize its sequencer by launching a first-of-its-kind sequencer pool. Opening up sequencer duties to many independent parties would eliminate the single-point-of-failure risk that a single sequencer entails.

By decentralizing sequencer duties across a much broader network, Metis will move beyond the industry standard of relying on a single sequencer entity, ensuring continuity of operations, robust network performance, and enhanced security. It will propel Metis closer to achieving the three challenges of what Ethereum co-founder Vitalik Buterin called the Blockchain Trilemma: decentralization, security, and scalability.

The sequencer pool will also bring new opportunities to Metis users. It will enable single-staking opportunities for METIS previously unavailable on our Layer 2 network; to become eligible to be a sequencer, interested parties will need to stake METIS.

To make that sequencer pool and the opportunities that it will bring possible requires high levels of data availability and simplicity – higher than what is currently available under the existing off-chain storage layer model. Migrating to the Optimistic Rollup model would address both of those challenges.

Community Implications

This move does come with a drawback: transaction fees would go from being lower than on any other Layer 2 network thanks to the Metis Smart Layer 2 approach, to being in line with other L2s. The Metis community could, at a later date, propose that Metis return to its Smart L2 approach after we launch and successfully roll out the sequencer pool. Note that for any such move to be considered, additional steps would need to be implemented to ensure optimal levels of data availability and security.

The need to retain Optimistic Rollup architecture would drop once the sequencer pool is up and running, since the initial challenges of (primarily) simplicity and (secondarily) data availability required would have already passed. If Metis switches to Optimistic Rollup architecture and then retains that format, ​​the call data pushed to Memo Labs in the past would require an extra step to access. This would not impact on-chain transaction verification, user experience, or the smooth operation of the network.

However, a blockchain is typically designed to make records of all data easily accessible to anyone who wants to go back and review any network activity. The community could ask Metis to provide a snapshot enabling easier access to all historical call data, complete with timestamps and other key details. Then after Ethereum migrates to EIP-4844, the community could propose that Metis evaluate the developer time and cost required to upload historical call data from Memo Labs to Ethereum, and decide how to proceed from there.

For now, though, the proposal before the community is a simple YES or NO (or ABSTAIN) on Metis migrating to Optimistic Rollup format to support full data availability.

Main Advantages for converting to Optimistic Rollup format

– Removes a layer of complexity in Metis architecture, thus facilitating the successful launch of the sequencer pool.

– Enables full data availability, which would help the launch of the sequencer pool.

– Sequencer pool will enable Metis to stop relying on a single entity to handle sequencer duties for the entire network, enabling more decentralization and security upgrades.

– Sequencer pool will enable single-staking opportunities for METIS not currently available in the ecosystem.

Main Disadvantage for converting to Optimistic Rollup format

– Would move transaction costs from lowest among all Layer 2 networks to in line with other L2s, pending a possible switch back to Smart L2 format once the sequencer pool is operational.

Next Steps

  1. Gather community feedback on this proposal.
  2. Once consensus is reached, move this proposal to the Snapshot stage for a community vote.
  3. If the Snapshot outcome is 51% YES, implement the proposed changes to Metis’s Data Availability, migrating it back to Ethereum Mainnet

I vote YES to METIS converting to Optimistic Rollup.


Yes ser!!!

I support the move!


Hi Metis team!

Another option to keep the costs down and decentralization high is to use Syscoin’s PoDA for DA. We are working with the layer zero team to add ZK light client of Syscoin to Ethereum so you can simply check a message that DA exists, the costs will be offloaded to SYS which is secured through Bitcoin miners. So not only do you increase security, you increase cost efficiency while maintaining a familiar model to the one you were already running. (its keccak hash based DA so nothing needs to change on your end)

1 Like

Thanks for the thorough explanation of different aspects of the technology in the proposal.

I’ve got a couple of questions -

  1. Are there any other options other than moving Data Availability to Ethereum? I’ve been increasingly hearing more about the ‘modular blockchain’ and curious on whether solutions such as Celestia or even NEAR Protocol have been considered as alternative Data Availability solutions.

  2. Timing. Is there a preferred timeline for implementing this? Unless there are any overriding technical concerns, I believe that equal or greater weight should be given to adoption and user growth. If lower fees are a major differentiating factor right now, then killing it too soon would be net negative. Considering there is intense competition among L2s for market share, I’d be wary of moving ahead if we end up being less differentiated where it matters (user experience, cost).

  3. Backwards compatibility - would there be any issues or considerations for existing applications, developers or users?

  4. Shift back to off-chain DA. I would suggest, after considering all options, to bundle this proposal with a clear pathway towards reverting to the most optimal DA solution after the d sequencer has been deployed. I believe strongly that unless we can compete on fees, any engineering accomplishment are irrelevant.



1 Like

As long as gas stays low, i like


Hello Metis Team,

Decentralizing sequencer duties seems like a significant step towards enhancing security and network performance. How exactly will interested parties stake METIS to become eligible to be a sequencer in the pool?

1 Like

I vote yes.

There are numerous benefits to it. Firstly, it reduces transaction costs. Additionally, I appreciate the fact that it enables single staking opportunities.

1 questions
How secure is this?


The proposal to switch to Optimistic Rollup format for Metis brings significant benefits:

  1. Simplification and Sequencer Pool Launch: It streamlines operations, making way for the sequencer pool, a major advancement.
  2. Enhanced Data Availability: Ensures all transaction data is available, aligning with Ethereum’s principles.
  3. Decentralization and Security: Spreading sequencer duties boosts network security and decentralization.

However, there’s a trade-off with potentially higher transaction costs. Despite this, the overall advantages make the switch worthwhile. I’m in favor of the proposal.

1 Like

I am bothered about Ethereum’s high gas fee, but I worry not since Metis offer the lowest gas fee. Then it’s a good idea.

1 Like


  1. Can you provide more details on the technical aspects of the migration process, including any potential risks or challenges that the Metis team anticipates during the transition to Optimistic Rollup format?
  2. How does migrating to the Optimistic Rollup format align with Metis’ long-term vision and goals for its Layer 2 network?
1 Like

Shifting from off-chain storage to the original Optimistic Rollup format on Layer 1 Ethereum is a strategic move by Metis, potentially enhancing transaction efficiency and scalability while aligning with Ethereum’s fundamental design.

1 Like

Hi blesily6688,

We will have more information released along with the testing :slight_smile:

1 Like


I’m in support of Metis migrating to Optimistic Rollup format. This will foster data availability and also bring new opportunities to Metis users.

1 Like

It sounds interesting, but I would like to know more about the technical aspects of it and how much security it has?
And more extensive explanations about single-staking opportunities for METIS!

1 Like

The advantages of this change outweigh the disadvantages. The Metis team is committed to data availability and decentralization, and this change will help them to achieve those goals. The sequencer pool will also be a valuable addition to the Metis ecosystem, and the single-staking opportunities for METIS will be attractive to users.

1 Like

How does decentralizing sequencer duties across a broader network in Metis enhance network performance and security while minimizing the risk of a single point of failure?

1 Like

Hello Metis!

  • What is the main motivation for Metis to migrate to an Optimistic Rollup architecture?
  • What are the benefits of decentralizing the sequencer?
  • What are the drawbacks of moving to an Optimistic Rollup model?
  • How will the community be involved in the decision-making process?
  • What are the next steps for Metis?
1 Like

Hello AVB,

  1. All other solutions will add the complexity of the DA solutions, for now, our priority is to keep Sequencer launching and operating successfully

  2. It will be changed together with the d-sequencer pool launch. Based on the past operation history, low fees didn’t bring extra transactions.

  3. Existing dapps do not need to change anything, whoever, developers who run the replica node will need to upgrade to the latest version later once the upgrade is done. (nothing needs to be done now

  4. When the d-sequencer is operated normally, we can revisit this topic and have a more concrete plan for migrating or not.

1 Like

Hi Patoae,

The sequencer nodes will be whitelisted at the first stage to guarantee the safe transition. Interested parties can speak with our BD team to know more about the details and requirements. Also, there will be some 3rd party DApps launching for people to stake Metis to participate in the sequencer nodes operation.

1 Like