Slot Leader Docker

Expanding on papacarp's pooltool.io repo and specifically on the leaderLogs scripts, this repo provides a dockerized version of leaderLogs.py intended for programmatic use.

NOTE: This currently only supports the mainnet!

Usage

To just run slot-leader once run:

docker run -ti --rm \
        -e EPOCH=252 \
	-e POOL_ID=079c...YOUR_POOL_ID_HEX...2791 \
	-v /path/to/your/ledger.json:/ledger.json \
	-v /path/to/your/vrf.skey:/vrf.skey \
	spectrumpool/slot-leader

This will download current epoch info from epoch-api (see below on how to circumvent), calculate σ and play through the epoch checking if your pool is slot leader for each slot.

To use it programmatically make sure to set the SILENT variable to "true"

docker run -ti --rm \
	-e SILENT=true \
        -e EPOCH=252 \
	-e POOL_ID=079c374160b0ae34a5a20b8e95a5b5c8766239b2984f13d7ab962791 \
	-v /path/to/your/ledger.json:/ledger.json \
	-v /path/to/your/vrf.skey:/vrf.skey \
	spectrumpool/slot-leader

Output

[
	{
		"timestamp": "2020-10-07 16:36:02", 
		"timestamp_unix": 1602088562.0, 
		"stolen": false, 
		"epochSlot": 406271
	}
]

The main output of the spectrumpool/slot-leader container is a JSON list of slots for which your pool is slot leader. The slot objects include a human readable timestamp, a unix timestamp and a boolean field indicating if this slot gets stolen by the legacy BFT protocol.

Without the SILENT flag some human oriented output is produced. In silent mode only the list of slots for which your pool will be slotleader will be printed out.

Other Input

You can pass an explicit SIGMA value which will circumvent the σ calculation.

docker run -ti --rm \
	-e SILENT=true \
        -e EPOCH=252 \
	-e SIGMA=0.0000197962 \
	-e POOL_ID=079c374160b0ae34a5a20b8e95a5b5c8766239b2984f13d7ab962791 \
	-v /path/to/your/ledger.json:/ledger.json \
	-v /path/to/your/vrf.skey:/vrf.skey \
	spectrumpool/slot-leader

Epochs, Epoch API dependency & Extra Arguments

You can steer wich epoch gets calculated via the EPOCH and EPOCH_NONCE env varliables. Since SIGMA can only be calculated for the current epoch, it is required to add these three env vars together.

This container is per default dependent on api.crypto2099.io for the epoch nonces. In case you add the EPOCH and EPOCH_NONCE env vars manually, the call to epoch-api is completely circumvented.

docker run -ti --rm \
	-e SILENT=true \
	-e SIGMA=0.0000065 \
	-e POOL_ID=079c374160b0ae34a5a20b8e95a5b5c8766239b2984f13d7ab962791 \
        -e EPOCH=252 \
        -e EXTRA_ARGS="--epoch-nonce 5ee77854fe91cc243b8d5589de3192e795f162097dba7501f8d1b0d5d7546bd5 --d-param 0.62" \
	-v /path/to/your/ledger.json:/ledger.json \
	-v /path/to/your/vrf.skey:/vrf.skey \
	spectrumpool/slot-leader

You can also pass arbitrary arguments to leaderLogs.py via the EXTRA_ARGS variable. See the begining of the leaderLogs.py file for what arguments can be passed.

Real World Compose Config

I run this on the start of each epoch via a systemd timer and service. Take a look at KIND Article (TBD) for more details on this.

services:
  slot-leader:
    image: spectrumpool/slot-leader
    environment:
      POOL_ID: 079c374160b0ae34a5a20b8e95a5b5c8766239b2984f13d7ab962791
    volumes:
    - /var/lib/cn/run/ledger.json:/ledger.json
    - /var/lib/vrf.skey:/vrf.skey

I don't trust your $%!@# container!

And that is fine. Just git clone https://github.com/SpectrumPool/slot-leader-docker.git and build it yourself! ;-)

Thanks

Thanks to Andrew Westberg [BCSH] and Papacarp [LOVE] for creating and hosting the meat of this container and to Adam Dean [BUFFY] for the nonce api.