diff --git a/dashboards/resources/workload.libsonnet b/dashboards/resources/workload.libsonnet index a2581cc3d..b31c204f8 100644 --- a/dashboards/resources/workload.libsonnet +++ b/dashboards/resources/workload.libsonnet @@ -1,324 +1,597 @@ -local grafana = import 'github.com/grafana/grafonnet-lib/grafonnet/grafana.libsonnet'; -local g = import 'github.com/grafana/jsonnet-libs/grafana-builder/grafana.libsonnet'; -local template = grafana.template; +local g = import 'github.com/grafana/grafonnet/gen/grafonnet-latest/main.libsonnet'; +local prometheus = g.query.prometheus; +local stat = g.panel.stat; +local table = g.panel.table; +local timeSeries = g.panel.timeSeries; +local var = g.dashboard.variable; { + local statPanel(title, unit, query) = + stat.new(title) + + stat.options.withColorMode('none') + + stat.standardOptions.withUnit(unit) + + stat.queryOptions.withInterval($._config.grafanaK8s.minimumTimeInterval) + + stat.queryOptions.withTargets([ + prometheus.new('${datasource}', query) + + prometheus.withInstant(true), + ]), + + local tsPanel = + timeSeries { + new(title): + timeSeries.new(title) + + timeSeries.options.legend.withShowLegend() + + timeSeries.options.legend.withAsTable() + + timeSeries.options.legend.withDisplayMode('table') + + timeSeries.options.legend.withPlacement('right') + + timeSeries.options.legend.withCalcs(['lastNotNull']) + + timeSeries.options.tooltip.withMode('single') + + timeSeries.fieldConfig.defaults.custom.withShowPoints('never') + + timeSeries.fieldConfig.defaults.custom.withFillOpacity(10) + + timeSeries.fieldConfig.defaults.custom.withSpanNulls(true) + + timeSeries.queryOptions.withInterval($._config.grafanaK8s.minimumTimeInterval), + }, + grafanaDashboards+:: { - local clusterTemplate = - template.new( - name='cluster', - datasource='$datasource', - query='label_values(up{%(kubeStateMetricsSelector)s}, %(clusterLabel)s)' % $._config, - current='', - hide=if $._config.showMultiCluster then '' else '2', - refresh=2, - includeAll=false, - sort=1 - ), - - local namespaceTemplate = - template.new( - name='namespace', - datasource='$datasource', - query='label_values(kube_namespace_status_phase{%(kubeStateMetricsSelector)s, %(clusterLabel)s="$cluster"}, namespace)' % $._config, - current='', - hide='', - refresh=2, - includeAll=false, - multi=false, - sort=1 - ), - - local workloadTypeTemplate = - template.new( - name='type', - datasource='$datasource', - query='label_values(namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace"}, workload_type)' % $._config.clusterLabel, - current='', - hide='', - refresh=2, - includeAll=true, - sort=0 - ), - - local workloadTemplate = - template.new( - name='workload', - datasource='$datasource', - query='label_values(namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload_type=~"$type"}, workload)' % $._config.clusterLabel, - current='', - hide='', - refresh=2, - includeAll=false, - sort=1 - ), 'k8s-resources-workload.json': - local tableStyles = { + local variables = { + datasource: + var.datasource.new('datasource', 'prometheus') + + var.datasource.withRegex($._config.datasourceFilterRegex) + + var.datasource.generalOptions.showOnDashboard.withLabelAndValue() + + var.datasource.generalOptions.withLabel('Data source') + + { + current: { + selected: true, + text: $._config.datasourceName, + value: $._config.datasourceName, + }, + }, + + cluster: + var.query.new('cluster') + + var.query.withDatasourceFromVariable(self.datasource) + + var.query.queryTypes.withLabelValues( + $._config.clusterLabel, + 'up{%(kubeStateMetricsSelector)s}' % $._config, + ) + + var.query.generalOptions.withLabel('cluster') + + var.query.refresh.onTime() + + ( + if $._config.showMultiCluster + then var.query.generalOptions.showOnDashboard.withLabelAndValue() + else var.query.generalOptions.showOnDashboard.withNothing() + ) + + var.query.withSort(type='alphabetical'), + + namespace: + var.query.new('namespace') + + var.query.withDatasourceFromVariable(self.datasource) + + var.query.queryTypes.withLabelValues( + 'namespace', + 'kube_namespace_status_phase{%(kubeStateMetricsSelector)s, %(clusterLabel)s="$cluster"}' % $._config, + ) + + var.query.generalOptions.withLabel('namespace') + + var.query.refresh.onTime() + + var.query.generalOptions.showOnDashboard.withLabelAndValue() + + var.query.withSort(type='alphabetical'), + + workload_type: + var.query.new('type') + + var.query.selectionOptions.withIncludeAll() + + var.query.withDatasourceFromVariable(self.datasource) + + var.query.queryTypes.withLabelValues( + 'workload_type', + 'namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace"}' % $._config, + ) + + var.query.generalOptions.withLabel('workload_type') + + var.query.refresh.onTime() + + var.query.generalOptions.showOnDashboard.withLabelAndValue() + + var.query.withSort(type='alphabetical'), + + workload: + var.query.new('workload') + + var.query.withDatasourceFromVariable(self.datasource) + + var.query.queryTypes.withLabelValues( + 'workload', + 'namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload_type=~"$type"}' % $._config, + ) + + var.query.generalOptions.withLabel('workload') + + var.query.refresh.onTime() + + var.query.generalOptions.showOnDashboard.withLabelAndValue() + + var.query.withSort(type='alphabetical'), + }; + + local links = { pod: { - alias: 'Pod', - link: '%(prefix)s/d/%(uid)s/k8s-resources-pod?var-datasource=$datasource&var-cluster=$cluster&var-namespace=$namespace&var-pod=$__cell' % { prefix: $._config.grafanaK8s.linkPrefix, uid: std.md5('k8s-resources-pod.json') }, + title: 'Drill down to pods', + url: '%(prefix)s/d/%(uid)s/k8s-resources-pod?${datasource:queryparam}&var-cluster=$cluster&var-namespace=$namespace&var-pod=${__data.fields.Pod}' % { + uid: $._config.grafanaDashboardIDs['k8s-resources-pod.json'], + prefix: $._config.grafanaK8s.linkPrefix, + }, }, }; + local cpuUsageQuery = ||| + sum( + node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"} + * on(namespace,pod) + group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload="$workload", workload_type=~"$type"} + ) by (pod) + ||| % $._config; + + local cpuRequestsQuery = ||| + sum( + kube_pod_container_resource_requests{%(kubeStateMetricsSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace", resource="cpu"} + * on(namespace,pod) + group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload="$workload", workload_type=~"$type"} + ) by (pod) + ||| % $._config; + + local cpuLimitsQuery = std.strReplace(cpuRequestsQuery, 'requests', 'limits'); + + local memUsageQuery = ||| + sum( + container_memory_working_set_bytes{%(clusterLabel)s="$cluster", namespace="$namespace", container!="", image!=""} + * on(namespace,pod) + group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload="$workload", workload_type=~"$type"} + ) by (pod) + ||| % $._config; + local memRequestsQuery = std.strReplace(cpuRequestsQuery, 'cpu', 'memory'); + local memLimitsQuery = std.strReplace(cpuLimitsQuery, 'cpu', 'memory'); + local networkColumns = [ ||| - (sum(irate(container_network_receive_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) + (sum(rate(container_network_receive_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) * on (namespace,pod) group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) ||| % $._config, ||| - (sum(irate(container_network_transmit_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) + (sum(rate(container_network_transmit_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) * on (namespace,pod) group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) ||| % $._config, ||| - (sum(irate(container_network_receive_packets_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) + (sum(rate(container_network_receive_packets_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) * on (namespace,pod) group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) ||| % $._config, ||| - (sum(irate(container_network_transmit_packets_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) + (sum(rate(container_network_transmit_packets_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) * on (namespace,pod) group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) ||| % $._config, ||| - (sum(irate(container_network_receive_packets_dropped_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) + (sum(rate(container_network_receive_packets_dropped_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) * on (namespace,pod) group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) ||| % $._config, ||| - (sum(irate(container_network_transmit_packets_dropped_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) + (sum(rate(container_network_transmit_packets_dropped_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace"}[%(grafanaIntervalVar)s]) * on (namespace,pod) group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) ||| % $._config, ]; - local networkTableStyles = { - pod: { - alias: 'Pod', - link: '%(prefix)s/d/%(uid)s/k8s-resources-pod?var-datasource=$datasource&var-cluster=$cluster&var-namespace=$namespace&var-pod=$__cell' % { prefix: $._config.grafanaK8s.linkPrefix, uid: std.md5('k8s-resources-pod.json') }, - }, - 'Value #A': { - alias: 'Current Receive Bandwidth', - unit: 'Bps', - }, - 'Value #B': { - alias: 'Current Transmit Bandwidth', - unit: 'Bps', - }, - 'Value #C': { - alias: 'Rate of Received Packets', - unit: 'pps', - }, - 'Value #D': { - alias: 'Rate of Transmitted Packets', - unit: 'pps', - }, - 'Value #E': { - alias: 'Rate of Received Packets Dropped', - unit: 'pps', - }, - 'Value #F': { - alias: 'Rate of Transmitted Packets Dropped', - unit: 'pps', - }, - }; + local panels = [ + tsPanel.new('CPU Usage') + + tsPanel.gridPos.withW(24) + + tsPanel.queryOptions.withTargets([ + prometheus.new('${datasource}', cpuUsageQuery) + + prometheus.withLegendFormat('__auto'), + ]), + table.new('CPU Quota') + + table.gridPos.withW(24) + + table.queryOptions.withTargets([ + prometheus.new('${datasource}', cpuUsageQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', cpuRequestsQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', cpuUsageQuery + '/' + cpuRequestsQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', cpuLimitsQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', cpuUsageQuery + '/' + cpuLimitsQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + ]) + + table.queryOptions.withTransformations([ + table.queryOptions.transformation.withId('joinByField') + + table.queryOptions.transformation.withOptions({ + byField: 'pod', + mode: 'outer', + }), - local cpuUsageQuery = ||| - sum( - node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"} - * on(namespace,pod) - group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload="$workload", workload_type=~"$type"} - ) by (pod) - ||| % $._config; + table.queryOptions.transformation.withId('organize') + + table.queryOptions.transformation.withOptions({ + excludeByName: { + Time: true, + 'Time 1': true, + 'Time 2': true, + 'Time 3': true, + 'Time 4': true, + 'Time 5': true, + }, + indexByName: { + 'Time 1': 0, + 'Time 2': 1, + 'Time 3': 2, + 'Time 4': 3, + 'Time 5': 4, + pod: 5, + 'Value #A': 6, + 'Value #B': 7, + 'Value #C': 8, + 'Value #D': 9, + 'Value #E': 10, + }, + renameByName: { + pod: 'Pod', + 'Value #A': 'CPU Usage', + 'Value #B': 'CPU Requests', + 'Value #C': 'CPU Requests %', + 'Value #D': 'CPU Limits', + 'Value #E': 'CPU Limits %', + }, + }), + ]) - local cpuRequestsQuery = ||| - sum( - kube_pod_container_resource_requests{%(kubeStateMetricsSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace", resource="cpu"} - * on(namespace,pod) - group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload="$workload", workload_type=~"$type"} - ) by (pod) - ||| % $._config; + + table.standardOptions.withOverrides([ + { + matcher: { + id: 'byRegexp', + options: '/%/', + }, + properties: [ + { + id: 'unit', + value: 'percentunit', + }, + ], + }, + { + matcher: { + id: 'byName', + options: 'Pod', + }, + properties: [ + { + id: 'links', + value: [links.pod], + }, + ], + }, + ]), - local cpuLimitsQuery = std.strReplace(cpuRequestsQuery, 'requests', 'limits'); + tsPanel.new('Memory Usage') + + tsPanel.gridPos.withW(24) + + tsPanel.standardOptions.withUnit('bytes') + + tsPanel.queryOptions.withTargets([ + prometheus.new('${datasource}', memUsageQuery) + + prometheus.withLegendFormat('__auto'), + ]), - local memUsageQuery = ||| - sum( - container_memory_working_set_bytes{%(clusterLabel)s="$cluster", namespace="$namespace", container!="", image!=""} - * on(namespace,pod) - group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload="$workload", workload_type=~"$type"} - ) by (pod) - ||| % $._config; - local memRequestsQuery = std.strReplace(cpuRequestsQuery, 'cpu', 'memory'); - local memLimitsQuery = std.strReplace(cpuLimitsQuery, 'cpu', 'memory'); + table.new('Memory Quota') + + table.gridPos.withW(24) + + table.standardOptions.withUnit('bytes') + + table.queryOptions.withTargets([ + prometheus.new('${datasource}', memUsageQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', memRequestsQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', memUsageQuery + '/' + memRequestsQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', memLimitsQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', memUsageQuery + '/' + memLimitsQuery) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + ]) + + table.queryOptions.withTransformations([ + table.queryOptions.transformation.withId('joinByField') + + table.queryOptions.transformation.withOptions({ + byField: 'pod', + mode: 'outer', + }), - g.dashboard( - '%(dashboardNamePrefix)sCompute Resources / Workload' % $._config.grafanaK8s, - uid=($._config.grafanaDashboardIDs['k8s-resources-workload.json']), - datasource_regex=$._config.datasourceFilterRegex, - datasource=$._config.datasourceName, - ) - .addRow( - g.row('CPU Usage') - .addPanel( - g.panel('CPU Usage') + - g.queryPanel(cpuUsageQuery, '{{pod}}') + - g.stack, - ) - ) - .addRow( - g.row('CPU Quota') - .addPanel( - g.panel('CPU Quota') + - g.tablePanel([ - cpuUsageQuery, - cpuRequestsQuery, - cpuUsageQuery + '/' + cpuRequestsQuery, - cpuLimitsQuery, - cpuUsageQuery + '/' + cpuLimitsQuery, - ], tableStyles { - 'Value #A': { alias: 'CPU Usage' }, - 'Value #B': { alias: 'CPU Requests' }, - 'Value #C': { alias: 'CPU Requests %', unit: 'percentunit' }, - 'Value #D': { alias: 'CPU Limits' }, - 'Value #E': { alias: 'CPU Limits %', unit: 'percentunit' }, - }) - ) - ) - .addRow( - g.row('Memory Usage') - .addPanel( - g.panel('Memory Usage') + - g.queryPanel(memUsageQuery, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('bytes') }, - ) - ) - .addRow( - g.row('Memory Quota') - .addPanel( - g.panel('Memory Quota') + - g.tablePanel([ - memUsageQuery, - memRequestsQuery, - memUsageQuery + '/' + memRequestsQuery, - memLimitsQuery, - memUsageQuery + '/' + memLimitsQuery, - ], tableStyles { - 'Value #A': { alias: 'Memory Usage', unit: 'bytes' }, - 'Value #B': { alias: 'Memory Requests', unit: 'bytes' }, - 'Value #C': { alias: 'Memory Requests %', unit: 'percentunit' }, - 'Value #D': { alias: 'Memory Limits', unit: 'bytes' }, - 'Value #E': { alias: 'Memory Limits %', unit: 'percentunit' }, - }) - ) - ) - .addRow( - g.row('Current Network Usage') - .addPanel( - g.panel('Current Network Usage') + - g.tablePanel( - networkColumns, - networkTableStyles - ), - ) - ) - .addRow( - g.row('Bandwidth') - .addPanel( - g.panel('Receive Bandwidth') + - g.queryPanel(||| - (sum(irate(container_network_receive_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) - * on (namespace,pod) - group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) - ||| % $._config, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('Bps') }, - ) - .addPanel( - g.panel('Transmit Bandwidth') + - g.queryPanel(||| - (sum(irate(container_network_transmit_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) - * on (namespace,pod) - group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) - ||| % $._config, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('Bps') }, - ) - ) - .addRow( - g.row('Average Container Bandwidth by Pod') - .addPanel( - g.panel('Average Container Bandwidth by Pod: Received') + - g.queryPanel(||| - (avg(irate(container_network_receive_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) - * on (namespace,pod) - group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) - ||| % $._config, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('Bps') }, - ) - .addPanel( - g.panel('Average Container Bandwidth by Pod: Transmitted') + - g.queryPanel(||| - (avg(irate(container_network_transmit_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) - * on (namespace,pod) - group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) - ||| % $._config, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('Bps') }, - ) - ) - .addRow( - g.row('Rate of Packets') - .addPanel( - g.panel('Rate of Received Packets') + - g.queryPanel(||| - (sum(irate(container_network_receive_packets_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) - * on (namespace,pod) - group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) - ||| % $._config, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('pps') }, - ) - .addPanel( - g.panel('Rate of Transmitted Packets') + - g.queryPanel(||| - (sum(irate(container_network_transmit_packets_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) - * on (namespace,pod) - group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) - ||| % $._config, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('pps') }, - ) - ) - .addRow( - g.row('Rate of Packets Dropped') - .addPanel( - g.panel('Rate of Received Packets Dropped') + - g.queryPanel(||| - (sum(irate(container_network_receive_packets_dropped_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) - * on (namespace,pod) - group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) - ||| % $._config, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('pps') }, - ) - .addPanel( - g.panel('Rate of Transmitted Packets Dropped') + - g.queryPanel(||| - (sum(irate(container_network_transmit_packets_dropped_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) - * on (namespace,pod) - group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) - ||| % $._config, '{{pod}}') + - g.stack + - { yaxes: g.yaxes('pps') }, - ) - ) + { - templating+: { - list+: [clusterTemplate, namespaceTemplate, workloadTypeTemplate, workloadTemplate], - }, - }, + table.queryOptions.transformation.withId('organize') + + table.queryOptions.transformation.withOptions({ + excludeByName: { + Time: true, + 'Time 1': true, + 'Time 2': true, + 'Time 3': true, + 'Time 4': true, + 'Time 5': true, + }, + indexByName: { + 'Time 1': 0, + 'Time 2': 1, + 'Time 3': 2, + 'Time 4': 3, + 'Time 5': 4, + pod: 8, + 'Value #A': 9, + 'Value #B': 10, + 'Value #C': 11, + 'Value #D': 12, + 'Value #E': 13, + }, + renameByName: { + pod: 'Pod', + 'Value #A': 'Memory Usage', + 'Value #B': 'Memory Requests', + 'Value #C': 'Memory Requests %', + 'Value #D': 'Memory Limits', + 'Value #E': 'Memory Limits %', + }, + }), + ]) + + + table.standardOptions.withOverrides([ + { + matcher: { + id: 'byRegexp', + options: '/%/', + }, + properties: [ + { + id: 'unit', + value: 'percentunit', + }, + ], + }, + { + matcher: { + id: 'byName', + options: 'Pod', + }, + properties: [ + { + id: 'links', + value: [links.pod], + }, + ], + }, + ]), + + table.new('Current Network Usage') + + table.gridPos.withW(24) + + table.queryOptions.withTargets([ + prometheus.new('${datasource}', networkColumns[0]) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', networkColumns[1]) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', networkColumns[2]) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', networkColumns[3]) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', networkColumns[4]) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + prometheus.new('${datasource}', networkColumns[5]) + + prometheus.withInstant(true) + + prometheus.withFormat('table'), + ]) + + + table.queryOptions.withTransformations([ + table.queryOptions.transformation.withId('joinByField') + + table.queryOptions.transformation.withOptions({ + byField: 'pod', + mode: 'outer', + }), + + table.queryOptions.transformation.withId('organize') + + table.queryOptions.transformation.withOptions({ + excludeByName: { + Time: true, + 'Time 1': true, + 'Time 2': true, + 'Time 3': true, + 'Time 4': true, + 'Time 5': true, + 'Time 6': true, + }, + indexByName: { + 'Time 1': 0, + 'Time 2': 1, + 'Time 3': 2, + 'Time 4': 3, + 'Time 5': 4, + 'Time 6': 5, + pod: 6, + 'Value #A': 7, + 'Value #B': 8, + 'Value #C': 9, + 'Value #D': 10, + 'Value #E': 11, + 'Value #F': 12, + }, + renameByName: { + pod: 'Pod', + 'Value #A': 'Current Receive Bandwidth', + 'Value #B': 'Current Transmit Bandwidth', + 'Value #C': 'Rate of Received Packets', + 'Value #D': 'Rate of Transmitted Packets', + 'Value #E': 'Rate of Received Packets Dropped', + 'Value #F': 'Rate of Transmitted Packets Dropped', + }, + }), + ]) + + + table.standardOptions.withOverrides([ + { + matcher: { + id: 'byRegexp', + options: '/Bandwidth/', + }, + properties: [ + { + id: 'unit', + value: 'Bps', + }, + ], + }, + { + matcher: { + id: 'byRegexp', + options: '/Packets/', + }, + properties: [ + { + id: 'unit', + value: 'pps', + }, + ], + }, + { + matcher: { + id: 'byName', + options: 'Pod', + }, + properties: [ + { + id: 'links', + value: [links.pod], + }, + ], + }, + ]), + + tsPanel.new('Receive Bandwidth') + + tsPanel.standardOptions.withUnit('Bps') + + tsPanel.queryOptions.withTargets([ + prometheus.new( + '${datasource}', + ||| + (sum(rate(container_network_receive_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) + * on (namespace,pod) + group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) + ||| % $._config + ) + + prometheus.withLegendFormat('__auto'), + ]), + + tsPanel.new('Transmit Bandwidth') + + tsPanel.standardOptions.withUnit('Bps') + + tsPanel.queryOptions.withTargets([ + prometheus.new( + '${datasource}', + ||| + (sum(rate(container_network_transmit_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) + * on (namespace,pod) + group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) + ||| % $._config + ) + + prometheus.withLegendFormat('__auto'), + ]), + + tsPanel.new('Average Container Bandwidth by Pod: Received') + + tsPanel.standardOptions.withUnit('Bps') + + tsPanel.queryOptions.withTargets([ + prometheus.new( + '${datasource}', + ||| + (avg(rate(container_network_receive_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) + * on (namespace,pod) + group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) + ||| % $._config + ) + + prometheus.withLegendFormat('__auto'), + ]), + + tsPanel.new('Average Container Bandwidth by Pod: Transmitted') + + tsPanel.standardOptions.withUnit('Bps') + + tsPanel.queryOptions.withTargets([ + prometheus.new( + '${datasource}', + ||| + (avg(rate(container_network_transmit_bytes_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) + * on (namespace,pod) + group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) + ||| % $._config + ) + + prometheus.withLegendFormat('__auto'), + ]), + + tsPanel.new('Rate of Received Packets') + + tsPanel.standardOptions.withUnit('pps') + + tsPanel.queryOptions.withTargets([ + prometheus.new( + '${datasource}', + ||| + (sum(rate(container_network_receive_packets_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) + * on (namespace,pod) + group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) + ||| % $._config + ) + + prometheus.withLegendFormat('__auto'), + ]), + + tsPanel.new('Rate of Transmitted Packets') + + tsPanel.standardOptions.withUnit('pps') + + tsPanel.queryOptions.withTargets([ + prometheus.new( + '${datasource}', + ||| + (sum(rate(container_network_transmit_packets_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) + * on (namespace,pod) + group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) + ||| % $._config + ) + + prometheus.withLegendFormat('__auto'), + ]), + + tsPanel.new('Rate of Received Packets Dropped') + + tsPanel.standardOptions.withUnit('pps') + + tsPanel.queryOptions.withTargets([ + prometheus.new( + '${datasource}', + ||| + (sum(rate(container_network_receive_packets_dropped_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) + * on (namespace,pod) + group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) + ||| % $._config + ) + + prometheus.withLegendFormat('__auto'), + ]), + + tsPanel.new('Rate of Transmitted Packets Dropped') + + tsPanel.standardOptions.withUnit('pps') + + tsPanel.queryOptions.withTargets([ + prometheus.new( + '${datasource}', + ||| + (sum(rate(container_network_transmit_packets_dropped_total{%(cadvisorSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace"}[%(grafanaIntervalVar)s]) + * on (namespace,pod) + group_left(workload,workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", %(namespaceLabel)s="$namespace", workload=~"$workload", workload_type=~"$type"}) by (pod)) + ||| % $._config + ) + + prometheus.withLegendFormat('__auto'), + ]), + ]; + + g.dashboard.new('%(dashboardNamePrefix)sCompute Resources / Workload' % $._config.grafanaK8s) + + g.dashboard.withUid($._config.grafanaDashboardIDs['k8s-resources-workload.json']) + + g.dashboard.withTags($._config.grafanaK8s.dashboardTags) + + g.dashboard.withEditable(false) + + g.dashboard.time.withFrom('now-1h') + + g.dashboard.time.withTo('now') + + g.dashboard.withRefresh($._config.grafanaK8s.refresh) + + g.dashboard.withVariables([variables.datasource, variables.cluster, variables.namespace, variables.workload_type, variables.workload]) + + g.dashboard.withPanels(g.util.grid.wrapPanels(panels, panelWidth=12, panelHeight=7)), }, }