OCP Performance Dashboard with new grafonnet library

.gitignore

@@ -1,4 +1,5 @@
 templates/grafonnet-lib
+templates/vendor
 rendered
 tmp
 bin

Makefile

@@ -1,45 +1,64 @@
-JSONNET = https://github.com/google/jsonnet/releases/download/v0.17.0/jsonnet-bin-v0.17.0-linux.tar.gz
+JB = https://github.com/jsonnet-bundler/jsonnet-bundler/releases/latest/download/jb-linux-amd64
 BINDIR = bin
 TEMPLATESDIR = templates
 OUTPUTDIR = rendered
 ALLDIRS = $(BINDIR) $(OUTPUTDIR)
 SYNCER_IMG_TAG ?= quay.io/cloud-bulldozer/dittybopper-syncer:latest
 PLATFORM = linux/amd64,linux/arm64,linux/ppc64le,linux/s390x
 
-# Get all templates at $(TEMPLATESDIR)
-TEMPLATES := $(wildcard $(TEMPLATESDIR)/**/*.jsonnet)
+ifeq ($(filter v2,$(MAKECMDGOALS)),v2)
+    # Set variables and instructions for v2
+  TEMPLATES := $(wildcard $(TEMPLATESDIR)/**/*-v2.jsonnet)
+	LIBRARY_PATH := $(TEMPLATESDIR)/vendor
+	JSONNET := https://github.com/cloud-bulldozer/utils/releases/download/v0.0.0/jsonnet-bin-v0.20.0-linux.tar.gz
+else
+	# Get all templates at $(TEMPLATESDIR)
+	TEMPLATES := $(filter-out %-v2.jsonnet, $(wildcard $(TEMPLATESDIR)/**/*.jsonnet))
+	LIBRARY_PATH := $(TEMPLATESDIR)/grafonnet-lib
+	JSONNET := https://github.com/google/jsonnet/releases/download/v0.17.0/jsonnet-bin-v0.17.0-linux.tar.gz
+endif
 
 # Replace $(TEMPLATESDIR)/*.jsonnet by $(OUTPUTDIR)/*.json
 outputs := $(patsubst $(TEMPLATESDIR)/%.jsonnet, $(OUTPUTDIR)/%.json, $(TEMPLATES))
 
 all: deps format build
 
-deps: $(ALLDIRS) $(TEMPLATESDIR)/grafonnet-lib $(BINDIR)/jsonnet
+deps: $(ALLDIRS) $(BINDIR)/jsonnet $(LIBRARY_PATH)
 
 $(ALLDIRS):
 	mkdir -p $(ALLDIRS)
 
 format: deps
 	$(BINDIR)/jsonnetfmt -i $(TEMPLATES)
 
-build: deps $(TEMPLATESDIR)/grafonnet-lib $(outputs)
+build: deps $(LIBRARY_PATH) $(outputs)
 
 clean:
 	@echo "Cleaning up"
-	rm -rf $(ALLDIRS) $(TEMPLATESDIR)/grafonnet-lib
-
-$(TEMPLATESDIR)/grafonnet-lib:
-	git clone --depth 1 https://github.com/grafana/grafonnet-lib.git $(TEMPLATESDIR)/grafonnet-lib
+	rm -rf $(ALLDIRS) $(TEMPLATESDIR)/vendor $(TEMPLATESDIR)/grafonnet-lib
 
 $(BINDIR)/jsonnet:
 	@echo "Downloading jsonnet binary"
 	curl -s -L $(JSONNET) | tar xz -C $(BINDIR)
+	@echo "Downloading jb binary"
+	curl -s -L $(JB) -o $(BINDIR)/jb
+	chmod +x $(BINDIR)/jb
+
+$(TEMPLATESDIR)/grafonnet-lib:
+	git clone --depth 1 https://github.com/grafana/grafonnet-lib.git $(TEMPLATESDIR)/grafonnet-lib
+
+$(TEMPLATESDIR)/vendor:
+	@echo "Downloading vendor files"
+	cd $(TEMPLATESDIR) && ../$(BINDIR)/jb install && cd ../
 
 # Build each template and output to $(OUTPUTDIR)
 $(OUTPUTDIR)/%.json: $(TEMPLATESDIR)/%.jsonnet
 	@echo "Building template $<"
 	mkdir -p $(dir $@)
-	$(BINDIR)/jsonnet $< > $@
+	$(BINDIR)/jsonnet -J ./$(LIBRARY_PATH) $< > $@
+
+v2: all
+	@echo "Rendered the v2 dashboards with latest grafonnet library"
 
 build-syncer-image: build
 	podman build --platform=${PLATFORM} -f Dockerfile --manifest=${SYNCER_IMG_TAG} .

README.md

@@ -34,6 +34,28 @@ bin/jsonnet templates/ocp-performance.jsonnet > rendered/ocp-performance.json
 $ ls rendered
 ocp-ingress-controller.json ocp-performance.json
 ```
+Similarly for V2, the dashboards that are built using latest grafonnet library, use 
+```
+$ make v2
+mkdir -p bin rendered
+Downloading jsonnet binary
+curl -s -L https://github.com/cloud-bulldozer/utils/releases/download/v0.0.0/jsonnet-bin-v0.20.0-linux.tar.gz | tar xz -C bin
+Downloading jb binary
+curl -s -L https://github.com/jsonnet-bundler/jsonnet-bundler/releases/latest/download/jb-linux-amd64 -o bin/jb
+chmod +x bin/jb
+Downloading vendor files
+cd templates && ../bin/jb install && cd ../
+GET https://github.com/grafana/grafonnet/archive/f40876da40d787e9c288de0b547ac85597c781d9.tar.gz 200
+GET https://github.com/grafana/grafonnet/archive/f40876da40d787e9c288de0b547ac85597c781d9.tar.gz 200
+GET https://github.com/jsonnet-libs/docsonnet/archive/cc9df63eaca56f39e8e4e1ce192141333257b08d.tar.gz 200
+GET https://github.com/jsonnet-libs/xtd/archive/0256a910ac71f0f842696d7bca0bf01ea77eb654.tar.gz 200
+bin/jsonnetfmt -i templates/General/ocp-performance-v2.jsonnet
+Building template templates/General/ocp-performance-v2.jsonnet
+mkdir -p rendered/General/
+bin/jsonnet -J ./templates/vendor templates/General/ocp-performance-v2.jsonnet > rendered/General/ocp-performance-v2.json
+Rendered the v2 dashboards with latest grafonnet library
+```
+Rest all operations reamin same as before.
 
 In order to clean up the environment execute `make clean`.
 

assets/ocp-performance/panels.libsonnet

@@ -0,0 +1,52 @@
+local g = import 'github.com/grafana/grafonnet/gen/grafonnet-latest/main.libsonnet';
+
+{
+  timeSeries: {
+    local timeSeries = g.panel.timeSeries,
+    local fieldOverride = g.panel.timeSeries.fieldOverride,
+    local custom = timeSeries.fieldConfig.defaults.custom,
+    local options = timeSeries.options,
+
+    generic(title, unit, targets, gridPos):
+      timeSeries.new(title)
+      + timeSeries.queryOptions.withTargets(targets)
+      + timeSeries.datasource.withUid('$datasource')
+      + timeSeries.standardOptions.withUnit(unit)
+      + timeSeries.gridPos.withX(gridPos.x)
+      + timeSeries.gridPos.withY(gridPos.y)
+      + timeSeries.gridPos.withH(gridPos.h)
+      + timeSeries.gridPos.withW(gridPos.w)
+      + custom.withSpanNulls('false')
+      + options.tooltip.withMode('multi')
+      + options.tooltip.withSort('desc')
+      + options.legend.withDisplayMode('table'),
+
+    genericLegend(title, unit, targets, gridPos):
+      self.generic(title, unit, targets, gridPos)
+      + options.legend.withShowLegend(true)
+      + options.legend.withCalcs([
+        'mean',
+        'max',
+        'min',
+      ])
+      + options.legend.withSortBy('max')
+      + options.legend.withSortDesc(true)
+      + options.legend.withPlacement('bottom'),
+  },
+  stat: {
+    local stat = g.panel.stat,
+    local options = stat.options,
+
+    base(title, targets, gridPos):
+      stat.new(title)
+      + stat.datasource.withUid('$datasource')
+      + stat.queryOptions.withTargets(targets)
+      + stat.gridPos.withX(gridPos.x)
+      + stat.gridPos.withY(gridPos.y)
+      + stat.gridPos.withH(gridPos.h)
+      + stat.gridPos.withW(gridPos.w)
+      + options.reduceOptions.withCalcs([
+        'last',
+      ]),
+  },
+}

assets/ocp-performance/queries.libsonnet

@@ -0,0 +1,201 @@
+local g = import 'github.com/grafana/grafonnet/gen/grafonnet-latest/main.libsonnet';
+local variables = import './variables.libsonnet';
+
+local generateTimeSeriesQuery(query, legend) = [
+   local prometheusQuery = g.query.prometheus;
+   prometheusQuery.new(
+    '$' + variables.datasource.name,
+    query
+   )
+   + prometheusQuery.withFormat('time_series')
+   + prometheusQuery.withIntervalFactor(2)
+   + prometheusQuery.withLegendFormat(legend),
+];
+
+{
+  nodeMemory: {
+    query(nodeName): 
+        generateTimeSeriesQuery('node_memory_Active_bytes{instance=~"' + nodeName + '"}', 'Active')
+        + generateTimeSeriesQuery('node_memory_MemTotal_bytes{instance=~"' + nodeName + '"}', 'Total')
+        + generateTimeSeriesQuery('node_memory_Cached_bytes{instance=~"' + nodeName + '"} + node_memory_Buffers_bytes{instance=~"' + nodeName + '"}', 'Cached + Buffers')
+        + generateTimeSeriesQuery('node_memory_MemAvailable_bytes{instance=~"' + nodeName + '"}', 'Available')
+        + generateTimeSeriesQuery('(node_memory_MemTotal_bytes{instance=~"' + nodeName + '"} - (node_memory_MemFree_bytes{instance=~"' + nodeName + '"} + node_memory_Buffers_bytes{instance=~"' + nodeName + '"} +  node_memory_Cached_bytes{instance=~"' + nodeName + '"}))', 'Used')
+  },
+  nodeCPU: {
+    query(nodeName):
+        generateTimeSeriesQuery('sum by (instance, mode)(irate(node_cpu_seconds_total{instance=~"' + nodeName + '",job=~".*"}[$interval])) * 100', 'Busy {{mode}}')
+  },
+  diskThroughput: {
+    query(nodeName):
+        generateTimeSeriesQuery('rate(node_disk_read_bytes_total{device=~"$block_device",instance=~"' + nodeName + '"}[$interval])', '{{ device }} - read')
+        + generateTimeSeriesQuery('rate(node_disk_written_bytes_total{device=~"$block_device",instance=~"' + nodeName + '"}[$interval])', '{{ device }} - write')
+  },
+  diskIOPS: {
+    query(nodeName):
+        generateTimeSeriesQuery('rate(node_disk_reads_completed_total{device=~"$block_device",instance=~"' + nodeName + '"}[$interval])', '{{ device }} - read')
+        + generateTimeSeriesQuery('rate(node_disk_writes_completed_total{device=~"$block_device",instance=~"' + nodeName + '"}[$interval])', '{{ device }} - write')
+  },
+  networkUtilization: {
+    query(nodeName):
+        generateTimeSeriesQuery('rate(node_network_receive_bytes_total{instance=~"' + nodeName + '",device=~"$net_device"}[$interval]) * 8', '{{instance}} - {{device}} - RX')
+        + generateTimeSeriesQuery('rate(node_network_transmit_bytes_total{instance=~"' + nodeName + '",device=~"$net_device"}[$interval]) * 8', '{{instance}} - {{device}} - TX')
+  },
+  networkPackets: {
+    query(nodeName):
+        generateTimeSeriesQuery('rate(node_network_receive_packets_total{instance=~"' + nodeName + '",device=~"$net_device"}[$interval])', '{{instance}} - {{device}} - RX')
+        + generateTimeSeriesQuery('rate(node_network_transmit_packets_total{instance=~"' + nodeName + '",device=~"$net_device"}[$interval])', '{{instance}} - {{device}} - TX')
+  },
+  networkDrop: {
+    query(nodeName):
+        generateTimeSeriesQuery('topk(10, rate(node_network_receive_drop_total{instance=~"' + nodeName + '"}[$interval]))', 'rx-drop-{{ device }}')
+        + generateTimeSeriesQuery('topk(10,rate(node_network_transmit_drop_total{instance=~"' + nodeName + '"}[$interval]))', 'tx-drop-{{ device }}')
+  },
+  conntrackStats: {
+    query(nodeName):
+        generateTimeSeriesQuery('node_nf_conntrack_entries{instance=~"' + nodeName + '"}', 'conntrack_entries')
+        + generateTimeSeriesQuery('node_nf_conntrack_entries_limit{instance=~"' + nodeName + '"}', 'conntrack_limit')
+  },
+  top10ContainerCPU: {
+    query(nodeName):
+        generateTimeSeriesQuery('topk(10, sum(irate(container_cpu_usage_seconds_total{container!="POD",name!="",node=~"' + nodeName + '",namespace!="",namespace=~"$namespace"}[$interval])) by (pod,container,namespace,name,service) * 100)', '{{ pod }}: {{ container }}')
+  },
+  top10ContainerRSS: {
+    query(nodeName):
+        generateTimeSeriesQuery('topk(10, container_memory_rss{container!="POD",name!="",node=~"' + nodeName + '",namespace!="",namespace=~"$namespace"})', '{{ pod }}: {{ container }}')
+  },
+  containerWriteBytes: {
+    query(nodeName):
+        generateTimeSeriesQuery('sum(rate(container_fs_writes_bytes_total{device!~".+dm.+", node=~"' + nodeName + '", container!=""}[$interval])) by (device, container)', '{{ container }}: {{ device }}')
+  },
+  stackroxCPU: {
+    query():
+        generateTimeSeriesQuery('topk(25, sum(irate(container_cpu_usage_seconds_total{container!="POD",name!="",namespace!="",namespace=~"stackrox"}[$interval])) by (pod,container,namespace,name,service) * 100)', '{{ pod }}: {{ container }}')
+  },
+  stackroxMem: {
+    query():
+        generateTimeSeriesQuery('topk(25, container_memory_rss{container!="POD",name!="",namespace!="",namespace=~"stackrox"})', '{{ pod }}: {{ container }}')
+  },
+  ovnAnnotationLatency: {
+    query():
+        generateTimeSeriesQuery('histogram_quantile(0.99, sum(rate(ovnkube_master_pod_creation_latency_seconds_bucket[$interval])) by (pod,le)) > 0', '{{ pod }}')
+  },
+  ovnCNIAdd: {
+    query():
+        generateTimeSeriesQuery('histogram_quantile(0.99, sum(rate(ovnkube_node_cni_request_duration_seconds_bucket{command="ADD"}[$interval])) by (pod,le)) > 0', '{{ pod }}')
+  },
+  ovnCNIDel: {
+    query():
+        generateTimeSeriesQuery('histogram_quantile(0.99, sum(rate(ovnkube_node_cni_request_duration_seconds_bucket{command="DEL"}[$interval])) by (pod,le)) > 0', '{{ pod }}')
+  },
+  ovnKubeMasterCPU: {
+    query():
+        generateTimeSeriesQuery('irate(container_cpu_usage_seconds_total{pod=~"ovnkube-master.*",namespace="openshift-ovn-kubernetes",container!~"POD|"}[$interval])*100', '{{container}}-{{pod}}-{{node}}')
+  },
+  ovnKubeMasterMem: {
+    query():
+        generateTimeSeriesQuery('container_memory_rss{pod=~"ovnkube-master-.*",namespace="openshift-ovn-kubernetes",container!~"POD|"}', '{{container}}-{{pod}}-{{node}}')
+  },
+  topOvnControllerCPU: {
+    query():
+        generateTimeSeriesQuery('topk(10, irate(container_cpu_usage_seconds_total{pod=~"ovnkube-.*",namespace="openshift-ovn-kubernetes",container="ovn-controller"}[$interval])*100)', '{{node}}')
+  },
+  topOvnControllerMem: {
+    query():
+        generateTimeSeriesQuery('topk(10, sum(container_memory_rss{pod=~"ovnkube-node-.*",namespace="openshift-ovn-kubernetes",container="ovn-controller"}) by (node))', '{{node}}')
+  },
+  promReplMemUsage: {
+    query():
+        generateTimeSeriesQuery('sum(container_memory_rss{pod="prometheus-k8s-1",namespace!="",name!="",container="prometheus"}) by (pod)', '{{pod}}')
+        + generateTimeSeriesQuery('sum(container_memory_rss{pod="prometheus-k8s-0",namespace!="",name!="",container="prometheus"}) by (pod)', '{{pod}}')
+  },
+  kubeletCPU: {
+    query():
+        generateTimeSeriesQuery('topk(10,irate(process_cpu_seconds_total{service="kubelet",job="kubelet"}[$interval])*100)', 'kubelet - {{node}}')
+  },
+  crioCPU: {
+    query():
+        generateTimeSeriesQuery('topk(10,irate(process_cpu_seconds_total{service="kubelet",job="crio"}[$interval])*100)', 'crio - {{node}}')
+  },
+  kubeletMemory: {
+    query():
+        generateTimeSeriesQuery('topk(10,process_resident_memory_bytes{service="kubelet",job="kubelet"})', 'kubelet - {{node}}')
+  },
+  crioMemory: {
+    query():
+        generateTimeSeriesQuery('topk(10,process_resident_memory_bytes{service="kubelet",job="crio"})', 'crio - {{node}}')
+  },
+  crioINodes: {
+    query():
+        generateTimeSeriesQuery('(1 - node_filesystem_files_free{fstype!="",mountpoint="/run"} / node_filesystem_files{fstype!="",mountpoint="/run"}) * 100', '/var/run - {{instance}}')
+  },
+  currentNodeCount: {
+    query():
+        generateTimeSeriesQuery('sum(kube_node_info{})', 'Number of nodes')
+        + generateTimeSeriesQuery('sum(kube_node_status_condition{status="true"}) by (condition) > 0', 'Node: {{ condition }}')
+  },
+  currentNamespaceCount: {
+    query():
+        generateTimeSeriesQuery('sum(kube_namespace_status_phase) by (phase)', '{{ phase }}')
+  },
+  currentPodCount: {
+    query():
+        generateTimeSeriesQuery('sum(kube_pod_status_phase{}) by (phase) > 0', '{{ phase}} Pods')
+  },
+  nsCount: {
+    query():
+        generateTimeSeriesQuery('sum(kube_namespace_status_phase) by (phase) > 0', '{{ phase }} namespaces')
+  },
+  podCount: {
+    query():
+        generateTimeSeriesQuery('sum(kube_pod_status_phase{}) by (phase)', '{{phase}} pods')
+  },
+  secretCmCount: {
+    query():
+        generateTimeSeriesQuery('count(kube_secret_info{})', 'secrets')
+        + generateTimeSeriesQuery('count(kube_configmap_info{})', 'Configmaps')
+  },
+  deployCount: {
+    query():
+        generateTimeSeriesQuery('count(kube_deployment_labels{})', 'Deployments')
+  },
+  servicesCount: {
+    query():
+        generateTimeSeriesQuery('count(kube_service_info{})', 'Services')
+  },
+  routesCount: {
+    query():
+        generateTimeSeriesQuery('count(openshift_route_info{})', 'Routes')
+  },
+  alerts: {
+    query():
+        generateTimeSeriesQuery('topk(10,sum(ALERTS{severity!="none"}) by (alertname, severity))', '{{severity}}: {{alertname}}')
+  },
+  podDistribution: {
+    query():
+        generateTimeSeriesQuery('count(kube_pod_info{}) by (node)', '{{ node }}')
+  },
+  top10ContMem: {
+    query():
+        generateTimeSeriesQuery('topk(10, container_memory_rss{namespace!="",container!="POD",name!=""})', '{{ namespace }} - {{ name }}')
+  },
+  top10ContCPU: {
+    query():
+        generateTimeSeriesQuery('topk(10,irate(container_cpu_usage_seconds_total{namespace!="",container!="POD",name!=""}[$interval])*100)', '{{ namespace }} - {{ name }}')
+  },
+  goroutinesCount: {
+    query():
+        generateTimeSeriesQuery('topk(10, sum(go_goroutines{}) by (job,instance))', '{{ job }} - {{ instance }}')
+  },
+  clusterOperatorsOverview: {
+    query():
+        generateTimeSeriesQuery('sum by (condition)(cluster_operator_conditions{condition!=""})', '{{ condition }}')
+  },
+  clusterOperatorsInformation: {
+    query():
+        generateTimeSeriesQuery('cluster_operator_conditions{name!="",reason!=""}', '{{name}} - {{reason}}')
+  },
+  clusterOperatorsDegraded: {
+    query():
+        generateTimeSeriesQuery('cluster_operator_conditions{condition="Degraded",name!="",reason!=""}', '{{name}} - {{reason}}')
+  },
+}