Streamline cluster configuration

[tjsturos]

There are two parts to this.

Starting control process with minimal config for clusters

The aim is that a Node Runner can add to their cluster by the following config changes:

...
engine:
  ...
  cluster:
    enabled: true
    machineConfigs:
      - ipAddress: 127.0.0.1 # required
        dataWorkerProcessesCount: 127 # required
        dataWorkerBaseListenPort: 40100 # default = 40000
    - ipAddress: 192.168.0.110
      dataWorkerProcessesCount: 128
   # ... etc.

this will tell the node's control process to create clients for addresses /ip4//tcp/<dataWorkerBaseListenPort + index>

So, the above config will create clients for:

• 127.0.0.1:40100-40227
• 192.168.0.110:40000-40128

Starting a data worker with no config

To start workers, a script just needs to start the data-workers:

For 127.0.0.1 config

~/ceremonyclient/node/node-1.4.21.1-amd-linux --core 1 --core-base-listen-port 40100

For 182.168.0.110

~/ceremonyclient/node/node-1.4.21.1-amd-linux --core 1 // defaults listen port to 40000

This could be scripted, such that:

DEFAULT_CORES=$(nproc)
## could find if there are already 'node' processes running and subtract, but for simplicities sake
DATA_WORKER_COUNT="${1:-$DEFAULT_CORES}"
BASE_LISTEN_PORT=${2:-40000}

CORE=1
while CORE <= $DATA_WORKER_COUNT; do
  ~/ceremonyclient/node/node-1.4.21.1-amd-linux --core $CORE --core-base-listen-port $BASE_LISTEN_PORT
   CORE=$CORE + 1
done