Chart name and version, where you feel a lack of requested feature
chart: victoria-metrics-k8s-stack
version: v0.73.0
Is your feature request related to a problem? Please describe.
Recently I upgraded the helm chart and this alertrule failed.
As it turned out this is because I have .Values.defaultRules.additionalGroupByLabels set to [namespace]
And having namespace in the group by of this rule doesn't make sense. The reason why I have namespace in additionalGroupByLabels is, because I host a few dozens of victoriametrics clusters in this kubernetes cluster, and i want alerts per namespace. But additionalGroupByLabels applys to basically all the alert rules, including the kubernetes mixins.
Describe the solution you'd like
So what I would like is to have an extra group by variable that only applies to the victoriametrics rules. Or another idea on how to solve this use case.
Describe alternatives you've considered
The only thing that I can come up with is to deploy the helm chart twice, once with the victoriametrics rules disabled and empty additionalGroupByLabels. And once with everything disabled but the victoriametrics rules. See additional context below for an example.
However, this approach has a drawback: if you later add another component that is enabled by default, it will break.
Additional context
Something like that, but it is not though through yet (there are some trues at the end):
alertmanager:
enabled: false
ingress:
enabled: false
monzoTemplate:
enabled: false
route:
enabled: false
coreDns:
enabled: false
service:
enabled: false
defaultDashboards:
dashboards:
node-exporter-full:
enabled: false
victoriametrics-operator:
enabled: false
victoriametrics-vmalert:
enabled: false
enabled: false
grafanaOperator:
enabled: false
defaultDatasources:
grafanaOperator:
enabled: false
grafana:
enabled: false
ingress:
enabled: false
sidecar:
dashboards:
enabled: false
datasources:
enabled: false
vmScrape:
enabled: false
kube-state-metrics:
autosharding:
enabled: false
customResourceState:
enabled: false
enabled: false
kubeRBACProxy:
enabled: false
kubeconfig:
enabled: false
networkPolicy:
enabled: false
prometheus:
monitor:
enabled: false
scrapeconfig:
enabled: false
securityContext:
enabled: false
selfMonitor:
enabled: false
service:
ipDualStack:
enabled: false
startupProbe:
enabled: false
verticalPodAutoscaler:
enabled: false
vmScrape:
enabled: false
kubeApiServer:
enabled: false
kubeControllerManager:
enabled: false
service:
enabled: false
kubeDns:
enabled: false
service:
enabled: false
kubeEtcd:
enabled: false
service:
enabled: false
kubeProxy:
enabled: false
service:
enabled: false
kubeScheduler:
enabled: false
service:
enabled: false
kubelet:
enabled: false
vmScrapes:
cadvisor:
enabled: false
probes:
enabled: false
resources:
enabled: false
prometheus-node-exporter:
enabled: false
hostRootFsMount:
enabled: false
kubeRBACProxy:
enabled: false
tls:
enabled: false
networkPolicy:
enabled: false
prometheus:
monitor:
enabled: false
podMonitor:
enabled: false
service:
enabled: false
ipDualStack:
enabled: false
terminationMessageParams:
enabled: false
tlsSecret:
enabled: false
verticalPodAutoscaler:
enabled: false
vmScrape:
enabled: false
victoria-metrics-operator:
crds:
cleanup:
enabled: false
enabled: false
serviceMonitor:
enabled: false
vmagent:
enabled: true
ingress:
enabled: false
route:
enabled: false
vmalert:
enabled: true
ingress:
enabled: false
route:
enabled: false
vmauth:
enabled: true
vmcluster:
enabled: true
ingress:
insert:
enabled: false
select:
enabled: false
storage:
enabled: false
route:
insert:
enabled: false
select:
enabled: false
storage:
enabled: false
spec:
vminsert:
enabled: false
vmselect:
enabled: false
vmdistributed:
enabled: false
vmsingle:
enabled: true
ingress:
enabled: false
route:
enabled: false
defaultRules:
create: false
Chart name and version, where you feel a lack of requested feature
chart: victoria-metrics-k8s-stack
version: v0.73.0
Is your feature request related to a problem? Please describe.
Recently I upgraded the helm chart and this alertrule failed.
As it turned out this is because I have
.Values.defaultRules.additionalGroupByLabelsset to[namespace]And having namespace in the group by of this rule doesn't make sense. The reason why I have namespace in
additionalGroupByLabelsis, because I host a few dozens of victoriametrics clusters in this kubernetes cluster, and i want alerts per namespace. ButadditionalGroupByLabelsapplys to basically all the alert rules, including the kubernetes mixins.Describe the solution you'd like
So what I would like is to have an extra group by variable that only applies to the victoriametrics rules. Or another idea on how to solve this use case.
Describe alternatives you've considered
The only thing that I can come up with is to deploy the helm chart twice, once with the victoriametrics rules disabled and empty
additionalGroupByLabels. And once with everything disabled but the victoriametrics rules. See additional context below for an example.However, this approach has a drawback: if you later add another component that is enabled by default, it will break.
Additional context
Something like that, but it is not though through yet (there are some trues at the end):