feat(monitoring): After some first analysis add cpu requests

After checking the current CPU usage using the following metric:

```prometheus
(sum(rate(container_cpu_usage_seconds_total{image!="",job="kubelet",metrics_path="/metrics/cadvisor", pod="prometheus-kube-prometheus-stack-prometheus-0"}[1m])) by (namespace, pod)) * 1000
```

800m seems quite good quite fitting when looking at the CPU usage of the
past 2 weeks.

This should help to keep all workload under control and produce better
scheduling for everything.