High response times when querying Loki series having cluster-reader clusterrole and namespace access granted in RHOL

Issue

• As a platform developer with cluster-reader Role in OpenShift Container Platform, when querying the Loki series takes up to 23 seconds when using Loki via Explore or Grafana dashboard with Loki functionality.

• As an user granted cluster-reader clusterrole and namespace access to the Loki logs, when it's queried the Loki series, the time is high. The same behaviour is not observed when querying the application logs

  $ time logcli -o raw --tls-skip-verify --bearer-token="$(oc whoami -t)" --addr "https://${lokiroute}/api/logs/v1/application" series --analyze-labels  --retries=2 --since=6h '{log_type="application"}'
  [...]
  Total Streams:  251
  Unique Labels:  10
  
  Label Name                 Unique Values  Found In Streams
  k8s_pod_name               251            251
  kubernetes_pod_name        251            251
  k8s_namespace_name         202            251
  kubernetes_namespace_name  202            251
  kubernetes_host            3              251
  k8s_node_name              3              251
  kubernetes_container_name  1              251
  openshift_log_type         1              251
  log_type                   1              251
  k8s_container_name         1              251
  
  real    0m53.729s
  user    0m0.117s
  sys 0m0.031s