status-benchmarks

Controlbox is a pod (single container) that will run inside the cluster to run the python scripts that will control the scenario. The docker image needs to be build with the ssh keys of the person that will connect to the controlbox pod.

In order to do this, go to controlbox folder:

Substitute <your_github_handle> with the handle of the used.
- These keys will be later used to ssh in the pod.
Build the image.
Make sure to use that image in deployment/codex/controlbox_codex.yaml

There is a bit more of information in the controlbox/README.md file.

Deploy everything

deploy_codex_local.sh will deploy:

Anvil, SNT contracts, services and everything that can be needed for more complex scenarios
Controlbox pod
K8s services
Nwaku bootstrap node
Nwaku store node
Status-codex nodes (without being initialized)
- You can modify how many nodes you want to deploy in the replicas field in status-backend-relay-codex.yaml

Note that if you want to repeat the deployment, you just need to run cleanup_local.sh and deploy_codex_local.sh again. If the cluster still has anvil and controlbox, there are things you need to comment in deploy_codex_local.sh. Those deployments are only needed on a fresh cluster.

Upload code to controlbox and execute

Run ./prepare_and_launch.sh

Editing and rerunning process is much faster and easier if you create an ssh venv from your IDE. Pycharm licensed supports this. I guess this can also be done with vscode. The only thing you need to do is enable the port forward, and the IDE should let you transfer the files to the controlbox, and create the venv. I am explaining here the slow-safe approach just in case.

There are other tools that also makes this process faster and easier (K9s to check logs, do portforwarding, etc.) But again, didn't want to introduce too many new dependencies.