Ethereum Secures the Rollup. But Who Controls the Sequencer?
Most users never think about the sequencer.
They should.
A rollup sequencer is the layer that sits between your transaction and the final settlement on Ethereum.
It decides:
• Which transactions are included
• The order they execute in
• How quickly users receive confirmations
• When batches are posted to Ethereum
• Whether normal L2 activity continues during an operational failure
A centralized sequencer can make an L2 extremely fast and efficient. But it also creates concentration around four major risks:
• Censorship: Who can exclude your transaction?
• Liveness: What happens when the sequencer goes offline?
• MEV: Who benefits from transaction ordering?
• Preconfirmations: How trustworthy are promises of inclusion before Ethereum settlement?
We have already seen why this matters.
Linea, 2024: During the Velocore exploit, Linea paused its sequencer and censored attacker addresses to limit further losses.
Arbitrum, 2023: A surge of inscription activity overwhelmed sequencing capacity and caused a significant interruption.
Base, 2026: Back to back sequencer related failures showed that even highly used L2s can experience operational fragility.
None of this means centralized sequencing is automatically bad.
It means the trade off should be understood.
The industry is experimenting with alternatives:
• Based sequencing
• Shared sequencers
• Decentralized sequencer sets
• MEV auctions and distributed sequencing
“How fast is this L2?”
Also ask:
• Who can halt it?
• Who controls transaction ordering?
• How does forced inclusion work?
• Where does sequencer MEV go?
• What happens if the operator fails?
• Can users still exit safely?
Because infrastructure decisions eventually become product decisions.
The sequencer sits in the critical gap between clicking “Submit” and believing your transaction actually happened.
That makes sequencing more than infrastructure.
It is part of the product.
✍️MMYamusa
Most users never think about the sequencer.
They should.
A rollup sequencer is the layer that sits between your transaction and the final settlement on Ethereum.
It decides:
• Which transactions are included
• The order they execute in
• How quickly users receive confirmations
• When batches are posted to Ethereum
• Whether normal L2 activity continues during an operational failure
A centralized sequencer can make an L2 extremely fast and efficient. But it also creates concentration around four major risks:
• Censorship: Who can exclude your transaction?
• Liveness: What happens when the sequencer goes offline?
• MEV: Who benefits from transaction ordering?
• Preconfirmations: How trustworthy are promises of inclusion before Ethereum settlement?
We have already seen why this matters.
Linea, 2024: During the Velocore exploit, Linea paused its sequencer and censored attacker addresses to limit further losses.
Arbitrum, 2023: A surge of inscription activity overwhelmed sequencing capacity and caused a significant interruption.
Base, 2026: Back to back sequencer related failures showed that even highly used L2s can experience operational fragility.
None of this means centralized sequencing is automatically bad.
It means the trade off should be understood.
The industry is experimenting with alternatives:
• Based sequencing
• Shared sequencers
• Decentralized sequencer sets
• MEV auctions and distributed sequencing
“How fast is this L2?”
Also ask:
• Who can halt it?
• Who controls transaction ordering?
• How does forced inclusion work?
• Where does sequencer MEV go?
• What happens if the operator fails?
• Can users still exit safely?
Because infrastructure decisions eventually become product decisions.
The sequencer sits in the critical gap between clicking “Submit” and believing your transaction actually happened.
That makes sequencing more than infrastructure.
It is part of the product.
✍️MMYamusa