Holochain Example

Mass deploys Holochain nodes with:

codeserver
dagu & server
ssh
mycelium
yggdrasil

How to use

The tfrobot requires Threefold Grid mneumonics so export them before running the example. Depending on which network tfrobot is configured to, use corresponding. (default is main)

export TFGRID_MNEMONIC="jelly fork  ..."

Run the holochain_deployer.vsh script. This will deploy VMs. You can interact with the vms over:

ssh
codeserver at port 8080
dagu api & ui at port 8081

go to ipaddress on port 8080 for seeing the codeserver. e.g. http://[300:463c:9082:e2d6:413d:46e0:c6b1:3de2]:8080/

Credentials

The credentials for the codeserver and dagu server can be set using the following environment variables. The default values for each variable is set below.

CODE_SERVER_PASSWORD=password # password for code server
DAGU_BASICAUTH_USERNAME=admin # dagu ui and api username
DAGU_BASICAUTH_PASSWORD=password # dagu ui and api password

These environment variables can be passed to the vms using the following configuration when calling tfrobot.deploy()

tfrobot.DeployConfig{
	vms: [
		tfrobot.VMConfig{
			env_vars: {
				'CODE_SERVER_PASSWORD': 'mypass'
			}
        }
    ]
}

DAG Example

The dag_example.vsh script demonstrates how the dagu client can be used to send and run a DAG on the machine to scaffold and serve a holochain web app.

To run:

./dag_example.vsh <MACHINE_IP>

This sends and runs the DAG, which can be monitored from <MACHINE_IP>:8081

Once the DAG reaches its final step and the web app starts running, the UI can be accessed by forwarding the following ports:

8282: Holochain web app
8888: Conductors send hello interface (optional)
Two conductor ports which are outputted by the final step of the DAG

Once these ports are forwarded, the Holochain web app can be accessed at localhost:8282, and will display the holo network with two conductors.

Phase 2

each vm will start sshserver and a call backhome
call back home is message over mycelium to the originator
- download/install hero as part of the init or hero in flist
- 'hero callhome $comma_separate_ipv6_list_mycelium'
- the hero call home will keep on trying untill it got confirmation that it was received (for 24 h)
- in other words sends alive message to the TFRobot node who deployed the VM
- TFRobot node runs vscript to do deploy, this vscript in loop keeps on showing progress of deploy as well as polling over mycelium if the alive message arrived
installed mycelium

ideas for later

integrate with https://dagu.readthedocs.io/en/latest/web_interface.html

2.6 KiB Raw Blame History