staking-miner

Staking Miner

Substrate chains validators compute a basic solution for the NPoS election. The optimization of the solution is computing-intensive and can be delegated to the staking-miner. The staking-miner does not act as validator and focuses solely on the optimization of the solution.

The staking miner connects to a specified chain and keeps listening to new Signed phase of the pallet-election-provider-multi-phase in order to submit solutions to the NPoS election. When the correct time comes, it computes its solution and submit it to the chain. The default miner algorithm is sequential-phragmen] with a configurable number of balancing iterations that improve the score.

Running the staking-miner requires passing the seed of a funded account in order to pay the fees for the transactions that will be sent. The same account's balance is used to reserve deposits as well. The best solution in each round is rewarded. All correct solutions will get their bond back. Any invalid solution will lose their bond.

You can check the help with:

staking-miner --help

Building

You can build from the root of the Polkadot repository using:

cargo build --profile production --locked --package staking-miner --bin staking-miner

Docker

There are 2 options to build a staking-miner Docker image:

• injected binary: the binary is first built on a Linux host and then injected into a Docker base image. This method only works if you have a Linux host or access to a pre-built binary from a Linux host.
• multi-stage: the binary is entirely built within the multi-stage Docker image. There is no requirement on the host in terms of OS and the host does not even need to have any Rust toolchain installed.

Building the injected image

First build the binary as documented above. You may then inject the binary into a Docker base image: parity/base-bin (running the command from the root of the Polkadot repository):

TODO: UPDATE THAT
docker build -t staking-miner -f scripts/ci/dockerfiles/staking-miner/staking-miner_injected.Dockerfile target/release

Building the multi-stage image

Unlike the injected image that requires a Linux pre-built binary, this option does not requires a Linux host, nor Rust to be installed. The trade-off however is that it takes a little longer to build and this option is less ideal for CI tasks. You may build the multi-stage image the root of the Polkadot repository with:

TODO: UPDATE THAT
docker build -t staking-miner -f scripts/ci/dockerfiles/staking-miner/staking-miner_builder.Dockerfile .

Running

A Docker container, especially one holding one of your SEED should be kept as secure as possible. While it won't prevent a malicious actor to read your SEED if they gain access to your container, it is nonetheless recommended running this container in read-only mode:

# The following line starts with an extra space on purpose:
 SEED=0x1234...

docker run --rm -i \
    --name staking-miner \
    --read-only \
    -e RUST_LOG=info \
    -e SEED=$SEED \
    -e URI=wss://your-node:9944 \
    staking-miner dry-run

Test locally

Make sure you've built Polkadot, then:

1. cargo run -p polkadot --features fast-runtime -- --chain polkadot-dev --tmp --alice -lruntime=debug
2. cargo run -p staking-miner -- --uri ws://localhost:9944 monitor --seed-or-path //Alice phrag-mms