Repository navigation
use a different network than the default public ipfs network #3
Description
Summary
Ideally, we want to connect only to nodes that run some lazyledger node, too.
Problem Definition
IPFS nodes connect to a large set of nodes by default. That's cool but we want our nodes to only connect to a set of nodes that 1) load the LL ipld plugin and 2) run some other portion of the lazyledger software (light client, (full) node, archival node etc).
Proposal
- Provide an alternative set of bootstrap nodes. Ideally, we'd automatically deploy a node before each release and its address to the default config (this is good for testing purposes).
- Add a way to only connect to a subnet. Using a private network (swarm key) could be the first very simple approach. Ideally, though we would modify to which nodes the ll-core nodes connect to by using some form of different network ID / protocol version (?) than the public ipfs network uses.
For Admin Use
- Not duplicate issue
- Appropriate labels applied
- Appropriate contributors tagged
- Contributor assigned/self-assigned
Activity
liamsi commentedon Feb 1, 2021
To summarize / capture some input we got on Discord from @Wondertan:
If we want, we can still use the default bootsrap nodes and use them to find other peers that use a specific protocol. e.g. here is a simple example that implements a chat-app and uses the default bootstrap nodes to find other nodes to talk to: https://github.com/libp2p/go-libp2p-examples/tree/master/chat-with-mdns
liamsi commentedon Mar 20, 2021
Triaging this depends to some extend on the outcome of running more benchmarks: celestiaorg/ipld-plugin-experiments#10
If rerunning the experiments yields OKish results, we do not need to act upon this before mainnet.
liamsi commentedon Aug 16, 2021
Transferred this from https://github.com/celestiaorg/celestia-core to https://github.com/celestiaorg/celestia-node
IMO, we should reconsider if we want to run a full-blown IPFS node in #1
Wondertan commentedon Aug 17, 2021
Using the public IPFS network basically means sharing global DHT. This could be a case even without using a full-blown IPFS node.
The definition of done for the issue:
Wondertan commentedon Aug 17, 2021
Wrong button 😅
Wondertan commentedon Nov 29, 2021
@liamsi, I feel like this can be closed. WDYT?
Wondertan commentedon Nov 29, 2021
Closing as we are using a completely different network.
docs/adr: ADR #3 March 2022 Testnet Design for Celestia Node (#244)
feat(core): add GetBlockInfo to Fetcher (#3)
services/header: Refactor `HeaderService` to be responsible for broad…
services/header: Refactor `HeaderService` to be responsible for broad…
bump dagstore #3