I recently updated the Grafana and kube-prometheus-stack Helm chart, and now in my Grafana dashboards, I'm seeing __auto instead of the actual pod names in the legend. This is happening across all dashboards that are using Prometheus queries for Kubernetes metrics.
Here’s an example of one of the queries:
sum(rate(container_network_receive_bytes_total{cluster="$cluster", namespace="$namespace"}[$__rate_interval])) by (pod)
And the legend format is set to __auto.
How do I update the dashboards to display the correct pod names instead of __auto?
Is there a way to bulk update the legend formatting across all Grafana dashboards, or should I update each dashboard manually? Any other configuration steps I might be missing?
Environment: Grafana Version: 8.5.8 Any help would be appreciated!
Things I’ve tried:
Manually updating the legend in the query to {{pod}} works, but I want to apply this change across all dashboards efficiently.
Checked the Prometheus relabeling configuration to ensure pod names are being relabeled correctly, and everything seems fine there.
I recently updated the Grafana and kube-prometheus-stack Helm chart, and now in my Grafana dashboards, I'm seeing __auto instead of the actual pod names in the legend. This is happening across all dashboards that are using Prometheus queries for Kubernetes metrics.
Here’s an example of one of the queries:
sum(rate(container_network_receive_bytes_total{cluster="$cluster", namespace="$namespace"}[$__rate_interval])) by (pod)And the legend format is set to __auto.How do I update the dashboards to display the correct pod names instead of __auto?
Is there a way to bulk update the legend formatting across all Grafana dashboards, or should I update each dashboard manually? Any other configuration steps I might be missing?
Environment: Grafana Version: 8.5.8 Any help would be appreciated!
Things I’ve tried:
Manually updating the legend in the query to {{pod}} works, but I want to apply this change across all dashboards efficiently.
Checked the Prometheus relabeling configuration to ensure pod names are being relabeled correctly, and everything seems fine there.