What Happens If a Data Availability Committee Fails to Publish Blobs?

25 October 2026

If a Data Availability Committee (DAC) fails to publish blobs, the rollup that depends on it can become temporarily unusable, and users may be unable to withdraw funds until the data is restored or the committee is replaced. In the worst case, assets could be frozen indefinitely.

Data Availability Committees are a compromise used by some Layer 2 networks to reduce costs. Instead of posting all transaction data to Ethereum (which is expensive), they entrust a small group of nodes to store and serve the data on demand. This works well when the committee is honest and available. When it is not, the rollup’s security model breaks.

What blobs are and why they matter

Blobs are large chunks of transaction data that rollups produce. On Ethereum, blobs are posted to the consensus layer via EIP-4844 (proto-danksharding) and are available to anyone for a limited time. On rollups using a DAC, blobs are stored off-chain by the committee. The DAC promises to make them available to anyone who asks - especially to challengers in optimistic rollups or to users trying to withdraw.

Without the blob data, no one can reconstruct the state of the rollup. That means:

• New blocks cannot be verified. The sequencer might still produce blocks, but nodes cannot validate them.
• Withdrawals become impossible. To exit the rollup, users must prove their balance on L1. That proof requires the blob data that shows their transaction was included.
• Fraud proofs cannot be submitted. In an optimistic rollup, anyone can challenge a state root by providing a fraud proof. Without the blob, the state root is unverifiable.

Immediate Effects of a DAC Failure

When a DAC stops publishing blobs, the rollup’s operators usually notice within minutes. The typical sequence is:

1. Sequencer halts. Most rollups pause block production to prevent invalid state growth. Transactions stop confirming.
2. Bridge freezes. The L1 bridge contract may lock deposits, or allow them but queue them until data is available.
3. Users cannot withdraw. Any pending withdrawal requests sit in a queue. The challenge period (if optimistic) cannot start because no one can verify the state.
4. Price of the rollup’s token (if any) drops. Market uncertainty about asset accessibility often triggers selling.

How long does the failure last?

That depends on the cause.

• Temporary outage (hours to days). The committee might be having a technical issue - a node went down, a cloud provider failed, or a software bug appeared. If the DAC members are reachable, they can restore service. The rollup can resume from the last published state.
• Permanent failure (weeks to never). If committee members disappear, go bankrupt, or are attacked, the data may be lost. Recovery requires a social fork: the community agrees on a new DAC or moves the rollup to a different data availability layer (like Ethereum or Celestia). This is slow and contentious.

Can users recover their funds?

Yes, but only if the blob data is eventually published or if the rollup’s governance can force a state reconstruction.

• If the data reappears: The rollup resumes. Withdrawals process normally. Users may face delays of days to weeks, but no loss.
• If the data is permanently lost: The rollup is effectively dead. No one can prove ownership of funds on L1. The bridge contract holds the deposited ETH or tokens, but it cannot release them because it does not know who owns what. Recovery requires a hard fork of the rollup’s L1 contracts to set a new state root - essentially a manual bailout by the community. This has never happened at scale for a major rollup.

How to Check If a Rollup Uses a DAC

Before using a rollup, check its data availability model. L2Beat lists this for every project under "Data Availability." If it says "External" or "Committee," the rollup relies on a DAC. If it says "Ethereum" or "On-chain," blobs are posted to L1.

Rollups that use DACs include many "validium" chains and some early-stage optimistic rollups. Examples (check current status on L2Beat): - Arbitrum Nova (uses a DAC) - zkSync Era (uses a DAC for some data) - Immutable X (validium)

Mitigation options for users

If you are using a DAC-based rollup and the committee fails:

1. Do not panic. Most outages are resolved within days. Check the rollup’s status page or Discord.
2. Do not try to bridge while the system is halted. Deposits may queue or fail. Withdrawals are blocked anyway.
3. If you hold a governance token, vote for a data availability upgrade. Some rollups allow token holders to replace the DAC or switch to Ethereum-based DA.
4. Consider moving to a rollup with on-chain data availability for assets you cannot afford to lose. The trade-off is higher fees for stronger guarantees.

The Bigger Picture

DAC failures are rare but not theoretical. In 2023, a DAC node for the Metis rollup went offline for several hours, halting the chain. The data was recovered, but the incident highlighted the risk. Rollups that use Ethereum for DA (like Arbitrum One, Optimism, Base) do not have this failure mode - blobs are always available as long as Ethereum is running.

If you prioritize safety over cost, use a rollup with on-chain data availability. If you need low fees and can accept the committee risk, DAC-based rollups can be fine for small amounts - but never for your entire portfolio.

Back to layer 2s