HOW TO

Prerequisites

• pyinfra

In order to deploy to the right cloud_controller we need some information of the user. This information should be provided in the inventory.py. The user should add an additional elif statement with the respective username and IP informations.

Furthermore the user has to edit the configurations files for continuum (cloud_conf.cfg and edge_conf.cfg). The changes need to be made to the following keys:

• base_path
• middleIP

Deploying the project

cd ./DistribSys-ServerlessEdge-Lab/deployment/

Due to the automation provided by pyinfra setting up the monitoring and running various tests is as easy as:

pyinfra inventory.py deploy.py

Open Grafana on your localhost

To view the dashboards locally in your browser you need to forward port 3000 from the cloud_controller to your localhost.

We found the easiest way to do this is by configuring proxy jumping from your host to the respective node on the dss cluster.

You can find an example ssh-config and command down below which forwards port 3000 from node4 on the dss cluster directly to your local host.

Host dss-cluster
        Hostname <DSS_CLUSTER_HOSTNAME>
        User <DSS_CLUSTER_USERNAME>
        IdentityFile <PATH_TO_PRIVATE_KEY_DSS_CLUSTER>
        PreferredAuthentications publickey

Host dss-node
        Hostname <DSS_NODE_HOSTNAME>
        User <DSS_NODE_USERNAME>
        ProxyJump dss-cluster
        IdentityFile <PATH_TO_PRIVATE_KEY_DSS_NODE>
        PreferredAuthentications publickey

ssh -L 3000:<CLOUD_CONTROLLER_IP>:3000 -J dss-cluster dss-node