Practise Test

1. Dockerize: the idea is to create self sufficient image with gaiad binary and all required configs. And meanwhile allow container to accept just usual arguments for gaiad .

   ---

• make build (optional) Builds images from current Dockerfile. The image is also available throught docker hub docker pull oplakida/gaia:v17.2.0 to pull locally
• docker run -p 26656 oplakida/gaia:v17.2.0 pruned start will start node in quick sync mode. After downloading pruned snapshot of db gaiad will start without any additional flags. One may pass any additional flags after ...start
• It's possible to feed docker run -p 26656 oplakida/gaia:v17.2.0 just commands as to gaiad binary
• It's possible to change node name, chain-id and genesis with corresponding flags to docker run command: -e MONIKER=<>, -e CHAIN_ID=<>, -e GENESIS_URL=<>

2. K8S: using image above deploy statefulSet with persistent volume(keep in mind size of volume is set to 10Gi in this example, but in eral life you will need 600-1500Gi depending on node type). Deployment is done with helm Chart and utilize default config files for node which could be cusotmized by helm values.yaml. Default cmd for container is set to pruned start wich triggers init_pruned.sh script. That script will init a new node and start quicksync process.

---

• edit helm/values.yaml file
• make deploy will run helm install --name gaia -n gaia ./helm

3. Observabilities: One of the part of previous step is to alter config.tml file and turn on exposure of prometheus metrics from container. In order to utilize this endpoint there is ServiceMonitor resource in the helm release. There should be installed Prometheus operator in order to scrape them.

4. Script kiddies: init_pruned.sh

5. Script grown-ups: TODO

6. Terraform lovers unite: All required resource including ones requiered for terraform initialization are in terraform directory. One just need to run terraform init && terraform apply in order to deploy everything. Although there are some prerequisites:

• There should be set up AWS creds.
• (Optional) Commmit out S3 remote backend in terraform/backend after first apply and apply once again if you want to keep remote state.