Linea DevOps/SRE/Platform Assignment

This a very simple example of Linea stack, without the ZeroKnowledge part. We'll use this simple docker compose to test out a candidates readiness to work with infrastructure at scale using Kubernetes and Helm.

Context

This docker compose consists of 4 parts:

• sequencer: an execution layer sequencer which is in charge of sequencing(ordering) transactions in a block
• maru: an in-house developed consensus client, similar to Teku and others, which is proposing and signing blocks (since we're running Paris fork)
• besu: a blockchain client rpc node receiving transactions and propagating them to the rest of the network
• ethstats: a web UI which displays network nodes status like latest block, number of peers, etc.

Task

The goal of this task is to convert this docker compose to production ready, publicly available Helm chart.
The following criteria should be met:

• Helm chart should be production ready, properly labeled and annotated.
• Helm chart should be available publicly.
• Helm chart should be deployable and the network fully working with default values and minimal user configuration.
• Helm chart repo should have a README.md file explaining its usage.
• All secrets must be handled properly.
• sequencer and maru are the core components of the network, and they should be treated as such from a security and reliability standpoint.
• Expose json-rpc and json-rpc-ws interfaces from besu node only, to the public.
• Expose ethstats service publicly.
• Implement ServiceMonitor CR which will scrape metrics and ship them into the pre-deployed observability stack.
• Implement basic autoscaling solution for besu RPC node, scaling on cpu and memory usage.
• Use a CronJob or a custom service which implements some form of data availability (backup) solution for sequencer, maru and besu nodes (data folder external backup, volume snapshot, etc.).