bacalhau
bacalhau copied to clipboard
Gateway Solution for (throttled) Networking
Opening this thread to discuss the goals for the upcoming Radius community grant related to researching solutions for a Bacalhau network gateway solution.
Goals:
- Enable some portion of traffic ingress/egress traffic between the Bacalhau network and the open internet. This capability would unlock many workloads and use cases for Bacalhau.
- Risk management of threats posed by open networking (e.g. botnet attacks).
Potential milestones:
- Research existing efforts malicious content and malicious behavior by other PLN technologies (IPFS, Filecoin, etc.)
- Research existing best practices among cloud providers to prevent use of their networks for botnet attacks (e.g. workload scanning, reputation, etc.).
- Create a proof of concept for a Bacalhau network gateway based on the research. This prototype could be in the form of code (PR), architecture diagrams, high level solution design, etc.
Open Questions:
- Which entity is best suited to determine which network traffic is approved and denied? PL, Bacalhau team, community moderation, etc.?
- What are the mechanisms by such approval would occur? Eg. the Akash (though only 'Experimental') has a central authority issue IP Leases (https://docs.akash.network/other-resources/experimental/ip-leases) and all traffic goes through a gateway. Are there other approaches?
- Are there benefits if the gateway solution is as decentralized, public, or trustless as the Filecoin network?