laconic-network-bootstrap/README.md

# laconic-network-bootstrap

A set of scripts to drive the mainnet-laconic stack stack orchestrator commands.

## Background

The mainnet-laconic stack implements a scheme for deploying multi-node networks where each node runs an instance of the stack.
The configuration of these nodes is done via a "network directory" created on each node, and by arranging to copy a genesis 
transaction file from each node to a coordinating node that generates a chain genesis file. This file in turn is copied
to the remaining nodes before stack startup. In addition, the IP addresses of all peer nodes need to be configured in 
each node. This scheme is implemented with the following stack orchestrator commands:

```
$ laconic-so --stack mainnet-laconic deploy setup --network-dir <network_dir> --initialize-network --chain-id <chain_id> --node-moniker <node_name>

$ laconic-so --stack mainnet-laconic deploy setup --network-dir <network_dir> --join-network --key-name <key_name>

$ laconic-so --stack mainnet-laconic deploy setup --network-dir <network_dir> --create-network --gentx-files <gentx_files>

$ laconic-so --stack mainnet-laconic deploy init --map-ports-to-host any-same --output <spec_file_name> --config "LACONIC_HOSTED_ENDPOINT=http://$<node_dns_name>:9473"

$ laconic-so --stack mainnet-laconic deploy create --deployment-dir <deployment_dir> --spec-file <spec_file_name> --network-dir <network_dir> --initial-peers <peers>
```

The scripts in this repository provide a convenient wrapper around these commands for the testnet use case
where there can be some level of coordination and common ownership between the nodes.

## Machine setup

The scripts assume a set of machines with dns host names assigned with the pattern <machine-name-prefix>-<node-number>.<machine-domain>, e.g.
testnet-a-1.testnets.servesthe.world
Testnet machines must be setup with stack orchestrator and docker installed.
The local machine (where the scripts are run) needs to be configured with `ssh-agent` and the relevant key added, such that ssh commands to connect to the testnet machines succeed without user interaction.

## Testnet deployment steps

```
$ cd scripts
```

In the example commands below we assume the machine-name-prefix "testnet-a".

### network.cfg

The scripts load a set of configuration variables from a file named `network.cfg`. An example file is provided, edit to suit:
```
$ cat network.cfg
machine_domain=testnets.servesthe.world
node_count=4
ssh_user=laconic
```

### Pull stack images

Run the 00-pull-stack-images.sh script to pull stack container images from a container registry:

```
$ ./01-pull-stack-images.sh testnet-a <image-registry-user> <image-registry-token>
```

### Initialize network directories

```
$ ./02-init-network-dirs.sh testnet-a
```

### Merge genesis files

```
$ ./03-merge-genesis-txns.sh testnet-a
```

Copy the peer list this script displays for use in the next step.

### Create deployments

```
$ ./04-create-deployments.sh testnet-a <peers-list>
```

### Start stacks

```
$ ./deployment-command testnet-a start
```

### Check testnet status

```
$ ./deployment-command testnet-a status
$ ./deployment-command testnet-a logs
```

The laconic console should be available on all nodes at http://<node-host-name>