Node Installation
1) Install AWS CLI tools
Since GoraNetwork uses a private Amazon ECR repository to host docker images, AWS CLI tools are necessary for local installs as well as cloud deployments. Refer to Amazon Getting started guide on how do download and install them.
2) Install GoraNetwork CLI tool
GoraNetwork CLI tool is available as a self-contained executable for Windows or Linux:
GoraNetwork CLI tool for Linux
GoraNetwork CLI tool for Windows
3) Configure GoraNetwork Algorand accounts
To ensure security of operator's staked assets, node Runner uses not one, but two Algorand accounts. Main account stakes operator's GORA tokens and receives all rewards earned by the node. Main account's private key is not exposed to the Node Runner. Participation account is auto-created by the Node Runner, then linked to the main account using GoraNetwork CLI tool. Participation account has no access to the operator's stake or rewards, except the small amount accrued since the last time rewards were claimed. This keeps staked tokens and most of rewards safe in case the Node Runner instance is compromised.
To create and initialize participation and main accounts use GoraNetwork Web app. This should leave you with an Algorand mnemonic and address for each account.
4) Create GoraNetwork configuration file
GoraNetwork configuration file contains essential information specific to the node. When deploying a node to the cloud or launching it locally, GoraNetwork CLI tool reads this file and passes its contents to the Node Runner via Docker container environment settings. The default location of this file is ~/.gora. Now create and fill it as follows:
Algorand node address must contain the protocol prefix, for example: "http://35.10.0.2". If the node uses a non-standard port, its number needs to be appended at the end after the usual ":".
After creating the file, set its access permissions to deny all access to anyone but yourself. On Linux, this is done by running: chmod 0600 ~/.gora.
5a) Deploy and run Node Runner on AWS
If you prefer to run your node locally, skip to section 5b.
The deployment step is only necessary for AWS. When performed for the first time, it creates the AWS objects necessary to run the Node Runner as a service: a Fargate task definition, an ECS cluster and an ECS service. Subsequent deployments only update these objects.
To perform a deployment, run gora deploy. If it is the first time, you should see an output like:
Warning! To prevent unintended service starts, the "deploy" command always leaves the service in a disabled state by setting "desiredCount" AWS parameter to 0, even if it was previously enabled (set to 1).
You must always enable the service after deployment by runnning gora enable. This kicks off the AWS service startup process which can sometimes take up to a few minutes. You can keep an eye on it by running gora status. Eventually, you should see an output like:
The IP address above is the public address of your GoraNetwork node. To check node status, point your browser to http://YOUR_NODE_IP:8000/ or inspect the logs using AWS CloudWatch logging service.
5b) Run Node Runner on any Docker-enabled host
When running a node for the first time or when updating it to the latest version, the Node Runner docker image needs to be pulled from GoraNetwork ECR private registry. Run gora docker-login to enable that. You should see an output like:
This step needs not be repeated on subsequent runs.
If your Algorand node runs on the same host, the Node Runner docker process may not be able to connect to its local address (127.0.0.1 or similar). To work around that, specify Docker machine IP instead, which can be found out by running ifconfig docker0 on Linux.
To start Node Runner, execute gora docker-start. When run for the first time, it will pull docker image from the registry. Then Node Runner will be launched, logging its messages to standard output, like:
5) Stop a Node Runner instance
To stop an AWS-deployed Node Runner instance, run gora disable. For stopping a locally run instance, hit Ctrl-C.
Last updated