Goal:
To make an endpoint we can use to download the entire metadata set of a site and mirror it or copy it,
Specification:
The beacon metadata bundler will have a simple job, which it will do continuously in a single sequence:
[ ] Analyse each entire OrbitDB database involved in serving its Riff site
[ ] Compile the data into a JSON bundle
[ ] Snapshot that bundle, upload that bundle to iroh to be served as a blob, upload that bundle to IPFS to be served
[ ] Get the CID of that blob
[ ] Publish the CID at an API endpoint (and if configured update an IPNS name via Kubo)
[ ] Bonus points if it keeps the set in memory at all times so that it can publish compressed snapshots very quickly.
The endpoint will be exposed only internally on an API path on port 5451, which will be used by a caching proxy that grabs the CIDs and publishes them for people to access (thus not allowing access directly to the beacon itself).
Goal: To make an endpoint we can use to download the entire metadata set of a site and mirror it or copy it,
Specification: The beacon metadata bundler will have a simple job, which it will do continuously in a single sequence:
The endpoint will be exposed only internally on an API path on port 5451, which will be used by a caching proxy that grabs the CIDs and publishes them for people to access (thus not allowing access directly to the beacon itself).