Data availability versus data storage: who archives rollup data post EIP-4844
Quote371 words
Distinguishes data availability from data storage. After EIP-4844, blobs are not stored permanently, so archival nodes run by indexers, explorers and apps are needed, and decentralized storage layers could be incentivized for this role.
Key claims
Calldata accounts for 80-90% of rollup cost today
Celestia verifies data availability but leaves historical storage to others
Blob expiry post EIP-4844 creates a need for archival nodes
0/ this also highlights the important difference between data availability and data storage
brief 🧵 (0/13)
x.com/hmalviya9/status/1660685561588244482?s=46&t=eiMbbnohgyPXY858ii-jqg
1/ data availability is concerned with whether the data published in
the latest block is available
2/ this is different from data storage, which is concerned with storing data securely and providing guarantees that it can be accessed when needed
3/ @CelestiaOrg is a L1 specialized on providing DA at scale. Hence, its purpose is to provide a secure and scalable way to verify the availability of data
want to know more? check here 👇
x.com/expctchaos/status/1535970188012572672
4/ once data has been verified as available, the job of storing and retrieving historical data is left up to other entities that require the
data
5/ luckily, there are natural incentives for outside parties to run what is referred to as archival nodes and store & serve historical data to users
6/ these actors might include block explorers, the indexers mentioned by @hmalviya9, but also applications & users themselves
7/ similarly, #EIP4844 will change how #Ethereum provides DA to rollups
with the introduction of blob carrying transactions, there will be data space (a.k.a. blobs of data) that rollups can post their data to, rather than having to pay for expensive calldata
8/ #EIP4844 is badly needed as calldata cost accounts for +80-90% of rollup cost today
for more info on EIP-4844, check my thread below 👇
x.com/expctchaos/status/1532100096111824901
9/ however, these data blobs are not stored permanently (unlike calldata) and hence, on Ethereum with state/blob expiry/pruning, we need archival nodes to ensure long term data storage, as this function is not covered by #Ethereum post #EIP4844 anymore
10/ but as mentioned earlier, runners of the rollups & indexers are naturally incentivised to run these archival nodes
11/ however, it would be nice to have actual incentivisation so we have more people permissionlessly running such nodes
12/ and while I don’t expect current rollup to integrate with storage layers, but rather fetch “old” data via the archival node providers as described above, it would be nice to see decentralized storage layers like @Filecoin or @ArweaveEco to be used in this context
13/ the thread below by @nickwh8te also explains the difference between data availability & storage nicely 👇
x.com/nickwh8te/status/1656321650475044865
QUOTING @hmalviya9
Data indexers will serve the massive data demand for L2/L3s. For example, DYDX is going to use indexers to store blockchain data. Additionally, after the implementation of EIP-4844, Ethereum L1 will also push data from Blobs to Indexers after a certain period.
Data indexers will serve the massive data demand for L2/L3s. For example, DYDX is going to use indexers to store blockchain data. Additionally, after the implementation of EIP-4844, Ethereum L1 will also push data from Blobs to Indexers after a certain period.
What came next · since 23 May 2023
Since then
Zero has written 31 more pieces on Celestia since this one.
- What If You Stop Seeing The Forest Because Of All The Trees?On Celestia
- Celestia 'Proof of Governance' proposal: 20x lower inflation and no stakingOn Celestia
- Why TIA is one of the most predatory VC tokens: pricing, unlocks and inflationOn Celestia
- Celestia vs EigenDA: different missions, not direct competitorsOn Celestia