I’m engaged on an on-chain knowledge mannequin in a company knowledge warehouse setting. All knowledge should stay on an inside system the place it may be simply joined to different datasources, and help quite a lot of excessive efficiency analytics use instances.
The mannequin will serve widespread stats you’d see in a block explorer like Mempool Area, nevertheless it must also help arbitrary queries like filtering by date or blockheight, or aggregations over some subset of addresses or deal with sorts. The key to quick combination question efficiency here’s a columnar database (i.e. Snowflake, Google BigQuery, or Apache HBase).
Whereas exploring choices to ingest on-chain knowledge, I discovered that every one of them require some type of Electrum Server, and plainly Electrs is most popular amongst trendy initiatives. What I’ve not been capable of finding is an in depth description of what precisely Electrum Server does on a technical degree. Does anybody know the place I can discover:
- A full technical description (not implementation) of all the pieces Electrum Server does to make knowledge extra out there to dam explorers and pockets software program? This can be a nice begin however solely covers the excessive degree thought: https://bitcoin.stackexchange.com/a/116582/142060
- Whether or not it is attainable to recreate this extra knowledge indexing, and so forth. from knowledge extracted immediately from Bitcoin Core? If not, how does one belief the outcomes that an Electrum Server supplies?
Thanks upfront