[DISCUSSION] rollups-node 2.0.0 helm chart

[endersonmaia]

Openning this issue so we can discuss possible designs for the rollups-node 2.0.0 chart.

Topics:

• It's a multiple app node instance
• A rollups-node can be instatiated and wait for app deployments
• We used to provide a container image with node and the app snapshot, we need a different strategy (see server-manager Pod to download a machine-snapshot from ipfs #28 )
• Single Pods vs Multiple Pods for individual services
• Which services should scale out (cartesi-rollups-jronrpc-api)
• Claimer should be isolated in a single Pod for better isolation?
• Claimer is the most critical one since it will handle mnemonic, aws kms and such, which security kubernetes provide for that?
• Ingress for inspect and jsonrpc
• Should we put a proxy over inspect and jsonrpc to have a single ingress ?
• Planning to use cartesi/rollups-runtime container image from cartes/cli repository
• Start simple with a Pod with everything inside in multiple containers (except for claimer, maybe) ?