flowchart TD
A[Branch Node] -->|Broadcast Signal| B(Leaf Node 1)
B -->|Mine| C(Submit to Stratum)
A -->|Broadcast Signal| D(Leaf Node 2)
D -->|Mine| E(Submit to Stratum)
A -->|Broadcast Signal| F(Leaf Node 3)
F -->|Mine| G(Submit to Stratum)
Cascade mining involves, a two actors implementation. A branch node and several leaf nodes (up to 3 in the optimal setup).
Each component is able to mine at full speed. For further details read the performances details in readme file.
Exposes, via mDNS, a broadcast signal to all the leaves.
At the ack of the connection with a leaf, the branch node provides:
- job identifier
- starting nonce value
- midstate hash
The branch then awaits for the job completion of each leaves and takes care to submit to stratum.
It's for now pretty simple implementation, according that nonces can run between 0 and 4,294,967,295 we start to split in halves.
So with a simple scheme:
| Branch | Leaf 1 |
|---|---|
| 0 - 2.147.483.647 | 2.147.483.648 - 4,294,967,295 |
| Branch | Leaf 1 | Leaf 2 |
|---|---|---|
| 0 - 1.431.655.765 | 1.431.655.766 - 2.863.311.530 | 2.863.311.531 - 4,294,967,295 |
Each leaf has basically 3 tasks:
- get a job
- calculate hash starting from midstate and incrementing nonce
- send back successful hashes to branch