[Dashboards] CoreDNS dashboard has a wrong job name

[sydorovdmytro]

Hello people,

CoreDNS dashboard is broken. The reason is the job name in queries, it's coredns:

"expr": "up{job=\"coredns\", instance=~\"$instance\"}" https://github.com/VictoriaMetrics/helm-charts/blob/master/charts/victoria-metrics-k8s-stack/templates/grafana/dashboards/k8s-system-coredns.yaml#L179 (just one example)

but it should be {{ printf "%s-%s" (include "victoria-metrics-k8s-stack.fullname" $) "coredns" | trunc 63 | trimSuffix "-" }}: https://github.com/VictoriaMetrics/helm-charts/blob/master/charts/victoria-metrics-k8s-stack/templates/servicemonitors/coredns.yaml#L30C9-L30C117

So the job name should not be hardcoded to a static value, or the ServiceMonitor needs a static name.

Another minor suggestion is to add a filter by the cluster label or at least add ad-hoc filters as we have for VM dashboards: https://github.com/VictoriaMetrics/helm-charts/blob/master/charts/victoria-metrics-k8s-stack/templates/grafana/dashboards/victoriametrics.yaml#L5611

Thanks