From ad0e83e27f37ad877696afd5a040bf9c91eb7f4b Mon Sep 17 00:00:00 2001 From: jeoffzheng Date: Thu, 2 Jul 2026 20:00:45 +0800 Subject: [PATCH 1/3] fix: replace humanizePercentage with humanize+% for 0-100 scale values KubeAggregatedAPIDown, KubernetesNodeHostHighCPUUsage, and KubernetesNodeHighDiskUsagePercentage all produce PromQL values on a 0-100 scale. humanizePercentage expects a 0-1 ratio and multiplies by 100, causing values like 81.82 to render as 8182% in notifications. Also fix typo: missing space before "has" in KubernetesNodeHostHighCPUUsage description. Bump chart version 1.3.0 -> 1.3.1 Bump plugindefinition spec.version 1.2.0 -> 1.2.1 and helmChart.version 1.3.0 -> 1.3.1 Signed-off-by: jeoffzheng --- charts/kubernetes-operations/Chart.yaml | 2 +- charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml | 2 +- charts/kubernetes-operations/alerts/kubernetes-node.yaml | 4 ++-- charts/kubernetes-operations/plugindefinition.yaml | 4 ++-- 4 files changed, 6 insertions(+), 6 deletions(-) diff --git a/charts/kubernetes-operations/Chart.yaml b/charts/kubernetes-operations/Chart.yaml index c4ac50e..0f5108c 100644 --- a/charts/kubernetes-operations/Chart.yaml +++ b/charts/kubernetes-operations/Chart.yaml @@ -3,7 +3,7 @@ apiVersion: v2 name: kubernetes-operations -version: 1.3.0 +version: 1.3.1 description: A set of Perses dashboards and Prometheus alerting rules combined with playbooks to ensure effective operations of Kubernetes. maintainers: - name: richardtief diff --git a/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml b/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml index ab7893e..91dfa42 100644 --- a/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml +++ b/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml @@ -53,6 +53,6 @@ groups: playbook: https://github.com/cloudoperators/kubernetes-operations/playbooks/KubeAggregatedAPIDown.md {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: - description: Kubernetes aggregated API `{{`{{ $labels.namespace }}`}}/{{`{{ $labels.name }}`}}` has been only `{{`{{ $value | humanizePercentage }}`}}` available over the last {{ dig "KubeAggregatedAPIDown" "for" "5m" .Values.prometheusRules }} . Run `kubectl get apiservice | grep -v Local` and confirm the services of aggregated APIs have active endpoints. + description: Kubernetes aggregated API `{{`{{ $labels.namespace }}`}}/{{`{{ $labels.name }}`}}` has been only `{{`{{ $value | humanize }}`}}%` available over the last {{ dig "KubeAggregatedAPIDown" "for" "5m" .Values.prometheusRules }} . Run `kubectl get apiservice | grep -v Local` and confirm the services of aggregated APIs have active endpoints. summary: Kubernetes aggregated API is down. {{- end }} diff --git a/charts/kubernetes-operations/alerts/kubernetes-node.yaml b/charts/kubernetes-operations/alerts/kubernetes-node.yaml index 3a4c398..81b8091 100644 --- a/charts/kubernetes-operations/alerts/kubernetes-node.yaml +++ b/charts/kubernetes-operations/alerts/kubernetes-node.yaml @@ -10,7 +10,7 @@ groups: playbook: https://github.com/cloudoperators/kubernetes-operations/playbooks/KubernetesNodeHostHighCPUUsage.md {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: - description: Node `{{`{{ $labels.node }}`}}`has more than `{{`{{ $value | humanizePercentage }}`}}` CPU load for {{ dig "KubernetesNodeHostHighCPUUsage" "for" "6h" .Values.prometheusRules }} + description: Node `{{`{{ $labels.node }}`}}` has more than `{{`{{ $value | humanize }}`}}%` CPU load for {{ dig "KubernetesNodeHostHighCPUUsage" "for" "6h" .Values.prometheusRules }} summary: High CPU load on node. {{- end }} @@ -75,7 +75,7 @@ groups: {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: description: Node disk usage above 85% - summary: "Disk usage on `{{`{{ $labels.node }}`}}` at `{{`{{ $value | humanizePercentage }}`}}`" + summary: "Disk usage on `{{`{{ $labels.node }}`}}` at `{{`{{ $value | humanize }}`}}%`" {{- end }} {{- if not (.Values.prometheusRules.disabled.KubernetesNodeNetworkUnavailable | default false) }} diff --git a/charts/kubernetes-operations/plugindefinition.yaml b/charts/kubernetes-operations/plugindefinition.yaml index 2f09942..e926da3 100644 --- a/charts/kubernetes-operations/plugindefinition.yaml +++ b/charts/kubernetes-operations/plugindefinition.yaml @@ -9,7 +9,7 @@ kind: PluginDefinition metadata: name: kubernetes-operations spec: - version: 1.2.0 + version: 1.2.1 displayName: Kubernetes operations bundle description: Operations bundle for Kubernetes docMarkDownUrl: https://raw.githubusercontent.com/cloudoperators/kubernetes-operations/main/README.md @@ -17,7 +17,7 @@ spec: helmChart: name: kubernetes-operations repository: oci://ghcr.io/cloudoperators/kubernetes-operations/charts - version: 1.3.0 + version: 1.3.1 options: - name: prometheusRules.create description: Create Prometheus rules From f092f149ca585093a53f2ced7222e5e963a0fa10 Mon Sep 17 00:00:00 2001 From: jeoffzheng Date: Mon, 6 Jul 2026 17:08:22 +0800 Subject: [PATCH 2/3] fix: use printf %.2f instead of humanize for percentage annotations Replace `humanize` with `printf "%.2f"` in all three alert descriptions to ensure consistent 2-decimal-place output. humanize can produce abbreviated values (1.1k, 1M) which is unsuitable for percentages. Also move the literal % outside the backtick-formatted value. Addresses review feedback from viennaa on PR #49. Signed-off-by: jeoffzheng --- charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml | 2 +- charts/kubernetes-operations/alerts/kubernetes-node.yaml | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml b/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml index 91dfa42..d80f2d9 100644 --- a/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml +++ b/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml @@ -53,6 +53,6 @@ groups: playbook: https://github.com/cloudoperators/kubernetes-operations/playbooks/KubeAggregatedAPIDown.md {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: - description: Kubernetes aggregated API `{{`{{ $labels.namespace }}`}}/{{`{{ $labels.name }}`}}` has been only `{{`{{ $value | humanize }}`}}%` available over the last {{ dig "KubeAggregatedAPIDown" "for" "5m" .Values.prometheusRules }} . Run `kubectl get apiservice | grep -v Local` and confirm the services of aggregated APIs have active endpoints. + description: Kubernetes aggregated API `{{`{{ $labels.namespace }}`}}/{{`{{ $labels.name }}`}}` has been only `{{`{{ printf "%.2f" $value }}`}}`% available over the last {{ dig "KubeAggregatedAPIDown" "for" "5m" .Values.prometheusRules }} . Run `kubectl get apiservice | grep -v Local` and confirm the services of aggregated APIs have active endpoints. summary: Kubernetes aggregated API is down. {{- end }} diff --git a/charts/kubernetes-operations/alerts/kubernetes-node.yaml b/charts/kubernetes-operations/alerts/kubernetes-node.yaml index 81b8091..10351cf 100644 --- a/charts/kubernetes-operations/alerts/kubernetes-node.yaml +++ b/charts/kubernetes-operations/alerts/kubernetes-node.yaml @@ -10,7 +10,7 @@ groups: playbook: https://github.com/cloudoperators/kubernetes-operations/playbooks/KubernetesNodeHostHighCPUUsage.md {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: - description: Node `{{`{{ $labels.node }}`}}` has more than `{{`{{ $value | humanize }}`}}%` CPU load for {{ dig "KubernetesNodeHostHighCPUUsage" "for" "6h" .Values.prometheusRules }} + description: Node `{{`{{ $labels.node }}`}}` has more than `{{`{{ printf "%.2f" $value }}`}}`% CPU load for {{ dig "KubernetesNodeHostHighCPUUsage" "for" "6h" .Values.prometheusRules }} summary: High CPU load on node. {{- end }} @@ -75,7 +75,7 @@ groups: {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: description: Node disk usage above 85% - summary: "Disk usage on `{{`{{ $labels.node }}`}}` at `{{`{{ $value | humanize }}`}}%`" + summary: "Disk usage on `{{`{{ $labels.node }}`}}` at `{{`{{ printf "%.2f" $value }}`}}`%" {{- end }} {{- if not (.Values.prometheusRules.disabled.KubernetesNodeNetworkUnavailable | default false) }} From e8f091301bf1ce1e00c9a7f5b56b58e70de0b227 Mon Sep 17 00:00:00 2001 From: jeoffzheng Date: Tue, 7 Jul 2026 14:39:08 +0800 Subject: [PATCH 3/3] fix: wrap description/summary in single quotes to avoid YAML parse error printf "%.2f" renders literal double-quotes into the YAML value. Unquoted YAML scalars and double-quoted strings both break when they contain unescaped double-quotes. Switch the three affected annotations to single-quote YAML scalars so the inner double-quotes are safe. Signed-off-by: jeoffzheng --- charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml | 2 +- charts/kubernetes-operations/alerts/kubernetes-node.yaml | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml b/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml index d80f2d9..35291df 100644 --- a/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml +++ b/charts/kubernetes-operations/alerts/kubernetes-apiserver.yaml @@ -53,6 +53,6 @@ groups: playbook: https://github.com/cloudoperators/kubernetes-operations/playbooks/KubeAggregatedAPIDown.md {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: - description: Kubernetes aggregated API `{{`{{ $labels.namespace }}`}}/{{`{{ $labels.name }}`}}` has been only `{{`{{ printf "%.2f" $value }}`}}`% available over the last {{ dig "KubeAggregatedAPIDown" "for" "5m" .Values.prometheusRules }} . Run `kubectl get apiservice | grep -v Local` and confirm the services of aggregated APIs have active endpoints. + description: 'Kubernetes aggregated API `{{`{{ $labels.namespace }}`}}/{{`{{ $labels.name }}`}}` has been only `{{`{{ printf "%.2f" $value }}`}}`% available over the last {{ dig "KubeAggregatedAPIDown" "for" "5m" .Values.prometheusRules }} . Run `kubectl get apiservice | grep -v Local` and confirm the services of aggregated APIs have active endpoints.' summary: Kubernetes aggregated API is down. {{- end }} diff --git a/charts/kubernetes-operations/alerts/kubernetes-node.yaml b/charts/kubernetes-operations/alerts/kubernetes-node.yaml index 10351cf..4760575 100644 --- a/charts/kubernetes-operations/alerts/kubernetes-node.yaml +++ b/charts/kubernetes-operations/alerts/kubernetes-node.yaml @@ -10,7 +10,7 @@ groups: playbook: https://github.com/cloudoperators/kubernetes-operations/playbooks/KubernetesNodeHostHighCPUUsage.md {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: - description: Node `{{`{{ $labels.node }}`}}` has more than `{{`{{ printf "%.2f" $value }}`}}`% CPU load for {{ dig "KubernetesNodeHostHighCPUUsage" "for" "6h" .Values.prometheusRules }} + description: 'Node `{{`{{ $labels.node }}`}}` has more than `{{`{{ printf "%.2f" $value }}`}}`% CPU load for {{ dig "KubernetesNodeHostHighCPUUsage" "for" "6h" .Values.prometheusRules }}' summary: High CPU load on node. {{- end }} @@ -75,7 +75,7 @@ groups: {{- include "kubernetes-operations.additionalRuleLabels" . | nindent 6 }} annotations: description: Node disk usage above 85% - summary: "Disk usage on `{{`{{ $labels.node }}`}}` at `{{`{{ printf "%.2f" $value }}`}}`%" + summary: 'Disk usage on `{{`{{ $labels.node }}`}}` at `{{`{{ printf "%.2f" $value }}`}}`%' {{- end }} {{- if not (.Values.prometheusRules.disabled.KubernetesNodeNetworkUnavailable | default false) }}