Polkadot Storage Phase 3
Beneficiary:Hey everyone, we are back with an improved long term plan for Polkadot Storage.
Last December, phase 3 got rejected during confirmation with a very narrow margin. Work has continued uninterrupted and we now have over 90% of phase 3 completed. We are testing the system in our private testnet and will move to a public testnet soon. Integration with mainnet has been postponed because of the JAM integration plans and we’ll continue improving tools and interfaces. For more details on our progress check the weekly updates on the forum.
During this time we have also been working with Parity and with their guidance have improved our long term approach. These ideas have been incorporated into Phase 4:
Jam Integration
- Using our storage for Jam DA resupply.
- Service Integration so that any Jam Service can use our long term storage.
Light weight proof system
- This was first initially flagged by Lucky Friday (thank you) and we later refined it during conversations with Parity. In short, our initial proof system protects against outsourcing attacks but it's also heavy and expensive. We wanna offer an alternative lighter proof system for users who don't need guarantees like that, at a much lower price.
- Merkle Proofs + zk-SNARKs
- Make the network work with multiple proof systems, let power users choose the proof system/guarantees they want.
Replication
- We added a naive version for now.
- Move from full replication to for example Erasure coding.
- Something like what Walrus uses “Twin-Code” or Reed Solomon error correction
Legal
- Explore responsibilities and liabilities when dealing with toxic data.
Economic model:
- Slashing - Although we have slashing implemented, we need to do some research on what the best numbers should be.
- Storage costs - Modeling the costs and income model for providers.
- Retrieval - Research retrieval costs to mitigate bandwidth attacks.
- Pricing - Price is important for users, make sure we do everything to be competitive with centralized solutions.
- Restaked DOT - Allow usage of restaked Dot so not to compete with staking rewards.
XCM
- Exposing the storage so that other parachains can store data. Check usefulness after Jam integration.
Marketplace model
- Thinking of this Storage layer also as a marketplace that gives access to all forms of competing centralized and decentralized storage.
Overall the plan is to turn this into a proper Storage product and build applications on top. With so much potential and attention on storage lately, it's an exciting time to build.
Links:
- Code https://github.com/eigerco/polka-storage/
- Book https://eigerco.github.io/polka-storage-book/
- Phase 1 https://polkadot.polkassembly.io/referenda/494
- Phase 2 https://polkadot.polkassembly.io/referenda/1150
- Old Phase 3 https://polkadot.polkassembly.io/referenda/1313
- Original proposal https://docs.google.com/document/d/16SZXvXyBRd-oEMIbgAid_vaxTuKzDlxuNc9CwSq3Lq0/edit
- Thinking about JAM DA integration https://docs.google.com/document/d/1L1jlMI9-KOfxMYUMQVeq0tPNtz1fitxTXlg6gKwjXnM/edit?usp=sharing
- Weekly updates https://forum.polkadot.network/t/polkadot-native-storage-updates/7021
- Research on native storage we first did in 2021 but there were limitations with collators back then. We returned to this topic in 2023 and got a grant from W3F to figure out how.
- Initial research pre-implementation https://github.com/eigerco/polkadot-native-storage/blob/main/doc/report/polkadot-native-storage-v1.0.0.pdf
- Old forum discussion https://forum.polkadot.network/t/polkadot-native-storage/4551
Comments (9)
Proposal Passed
3
of 3Summary
0%
Aye
0%
Nay
Aye (68)0.0 PAS
Support0.0 PAS
Nay (39)0.0 PAS
Comments (9)
Hi everyone,
we have discussed extensively with Eiger and we are pleased with the progress so far.
We are supporting funding this proposal "Phase 3" and will support "Phase 4". Phase 4 is the first milestones that propose a product more than a prototype and start to go into the fundamental question for storage: decentralized properties but also pricing (GB/month, IO/s month and QoS).
Why do we suport them?
- We need better storage on Polkadot. Storage is fundational like Compute. But without a strong storage solution, then you can only build product that can fit onchain and that's fine for DeFI but not for a lot of things.
- Eiger ways of working and commitment to Polkadot is great (no new token, deliver what they promised), Parity has also enough to do with the Compute part and is happy to have a strong partner that can "get it done".
- Robust storage model (decentralized with proof which you do not find everywhere)
Pierre
Note: storage is an expensive business. Building a distributed storage solution is hard and making it decentralized is harder. We expect to see a few milestones before this is a hardened and scalable solution. We plan to do it step by step and come back to you all each time. We also plan to get some prototype of product build on top to see how the system behave and to define a powerfull but easy to use interface. You should see some other proposals soonish for building on top of it services like a on chain password manager or a personal photo system proposed by other builders in the ecosystem.
Pierre from Parity
OG Tracker Rating 3/3
Clear display of deliverables✅
- polka-index
- polka-fetch
- Deployment
- Delia
- Gregor
Clear display of a valid direct point of contact✅
- Email: daren@eiger.co + hello@eiger.co
Clear display of proposal’s duration✅
- The duration of this proposal is 7 months.
OGT Rating aims to help voters make better informed decisions and direct proposers towards certain common-good practices. We are providing feedback based on 3 simple yet crucial criteria which we believe should be included in every OpenGov referenda.
Hi everyone,
we have discussed extensively with Eiger and we are pleased with the progress so far.
We are supporting funding this proposal "Phase 3" and will support "Phase 4". Phase 4 is the first milestones that propose a product more than a prototype and start to go into the fundamental question for storage: decentralized properties but also pricing (GB/month, IO/s month and QoS).
Why do we suport them?
- We need better storage on Polkadot. Storage is fundational like Compute. But without a strong storage solution, then you can only build product that can fit onchain and that's fine for DeFI but not for a lot of things.
- Eiger ways of working and commitment to Polkadot is great (no new token, deliver what they promised), Parity has also enough to do with the Compute part and is happy to have a strong partner that can "get it done".
- Robust storage model (decentralized with proof which you do not find everywhere)
Pierre
Note: storage is an expensive business. Building a distributed storage solution is hard and making it decentralized is harder. We expect to see a few milestones before this is a hardened and scalable solution. We plan to do it step by step and come back to you all each time. We also plan to get some prototype of product build on top to see how the system behave and to define a powerfull but easy to use interface. You should see some other proposals soonish for building on top of it services like a on chain password manager or a personal photo system proposed by other builders in the ecosystem.
Pierre from Parity
OG Tracker Rating 3/3
Clear display of deliverables✅
Clear display of a valid direct point of contact✅
Clear display of proposal’s duration✅
OGT Rating aims to help voters make better informed decisions and direct proposers towards certain common-good practices. We are providing feedback based on 3 simple yet crucial criteria which we believe should be included in every OpenGov referenda.