SigNoz is an open-source, OpenTelemetry-native observability platform for your team and their AI agents. Get logs, metrics, and traces in one tool with features like APM, distributed tracing, log management, infra monitoring, etc. Combined with SigNoz MCP and a native AI teammate (in SigNoz Cloud) it helps you build more resilient apps.
TypeScript apmapplication-monitoringdistributed-tracinggogood-first-issuejaegerloglogsmetricsmonitoringnextjsobservabilityopen-sourceopentelemetryprometheusreactreactjsself-hostedtracingtypescript
package implinframonitoring
import (
"github.com/SigNoz/signoz/pkg/types/inframonitoringtypes"
"github.com/SigNoz/signoz/pkg/types/metrictypes"
qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
)
var clusterNameGroupByKey = qbtypes.GroupByKey{
TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{
Name: inframonitoringtypes.ClusterNameAttrKey,
FieldContext: telemetrytypes.FieldContextResource,
FieldDataType: telemetrytypes.FieldDataTypeString,
},
}
// clustersTableMetricNamesList drives the existence/retention check.
// Includes k8s.node.condition_ready and k8s.pod.phase so the response
// short-circuits cleanly when a cluster doesn't ship those metrics — even
// though they aren't part of the QB composite query (they're queried separately
// via getPerGroupNodeConditionCounts and getPerGroupPodStatusCounts).
var clustersTableMetricNamesList = []string{
"k8s.node.cpu.usage",
"k8s.node.allocatable_cpu",
"k8s.node.memory.working_set",
"k8s.node.allocatable_memory",
"k8s.node.condition_ready", //TODO(nikhilmantri0902): should these metrics be used to count groups k8s.node.condition_ready and k8s.pod.phase
"k8s.pod.phase",
"k8s.pod.status_reason",
"k8s.container.status.reason",
}
var clusterAttrKeysForMetadata = []string{
"k8s.cluster.name",
}
// clusterCountAttrKeys are the resource attributes whose distinct values are
// counted per cluster. Node name is read from the node metric universe, while
// namespace + workload names come from the pod metric universe — both unioned
// into clusterCountMetricNamesList.
var clusterCountAttrKeys = []string{
inframonitoringtypes.NodeNameAttrKey,
inframonitoringtypes.NamespaceNameAttrKey,
inframonitoringtypes.DeploymentNameAttrKey,
inframonitoringtypes.DaemonSetNameAttrKey,
inframonitoringtypes.JobNameAttrKey,
inframonitoringtypes.StatefulSetNameAttrKey,
}
// clusterMetricNamesListForCounts is the metric universe for per-cluster distinct
// counts.
var clusterMetricNamesListForCounts = []string{
// for nodes lookup
"k8s.node.cpu.usage",
"k8s.node.allocatable_cpu",
"k8s.node.memory.working_set",
"k8s.node.allocatable_memory",
"k8s.node.condition_ready",
// for daemonsets lookup
"k8s.daemonset.desired_scheduled_nodes",
"k8s.daemonset.current_scheduled_nodes",
"k8s.daemonset.ready_nodes",
"k8s.daemonset.misscheduled_nodes",
// for deployments lookup
"k8s.deployment.desired",
"k8s.deployment.available",
// for jobs lookup
"k8s.job.active_pods",
"k8s.job.failed_pods",
"k8s.job.successful_pods",
"k8s.job.desired_successful_pods",
// for statefulsets lookup
"k8s.statefulset.desired_pods",
"k8s.statefulset.current_pods",
// for namespaces and general lookup for all
"k8s.pod.cpu.usage",
"k8s.pod.cpu_request_utilization",
"k8s.pod.cpu_limit_utilization",
"k8s.pod.memory.working_set",
"k8s.pod.memory_request_utilization",
"k8s.pod.memory_limit_utilization",
"k8s.pod.phase",
"k8s.pod.status_reason",
}
var orderByToClustersQueryNames = map[string][]string{
inframonitoringtypes.ClustersOrderByCPU: {"A"},
inframonitoringtypes.ClustersOrderByCPUAllocatable: {"B"},
inframonitoringtypes.ClustersOrderByMemory: {"C"},
inframonitoringtypes.ClustersOrderByMemoryAllocatable: {"D"},
}
// newClustersTableListQuery builds the composite QB v5 request for the clusters list.
// Cluster-scope metrics are derived by summing per-node metrics within the
// group (default group: k8s.cluster.name). Node condition counts and pod status
// counts are derived separately via getPerGroupNodeConditionCounts and
// getPerGroupPodStatusCounts respectively (works for both list and grouped_list
// modes), so neither is included here. Query letters A/B/C/D mirror the v1
// implementation and the v2 nodes list.
func (m *module) newClustersTableListQuery() *qbtypes.QueryRangeRequest {
queries := []qbtypes.QueryEnvelope{
// Query A: CPU usage — sum of node CPU within the group.
{
Type: qbtypes.QueryTypeBuilder,
Spec: qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation]{
Name: "A",
Signal: telemetrytypes.SignalMetrics,
Aggregations: []qbtypes.MetricAggregation{
{
MetricName: "k8s.node.cpu.usage",
TimeAggregation: metrictypes.TimeAggregationAvg,
SpaceAggregation: metrictypes.SpaceAggregationSum,
ReduceTo: qbtypes.ReduceToAvg,
},
},
GroupBy: []qbtypes.GroupByKey{clusterNameGroupByKey},
Disabled: false,
},
},
// Query B: CPU allocatable — sum of node allocatable CPU within the group.
// TimeAggregationLatest is the closest v5 equivalent of v1's AnyLast;
// allocatable values change rarely so divergence in practice is negligible.
{
Type: qbtypes.QueryTypeBuilder,
Spec: qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation]{
Name: "B",
Signal: telemetrytypes.SignalMetrics,
Aggregations: []qbtypes.MetricAggregation{
{
MetricName: "k8s.node.allocatable_cpu",
TimeAggregation: metrictypes.TimeAggregationLatest,
SpaceAggregation: metrictypes.SpaceAggregationSum,
ReduceTo: qbtypes.ReduceToAvg,
},
},
GroupBy: []qbtypes.GroupByKey{clusterNameGroupByKey},
Disabled: false,
},
},
// Query C: Memory working set — sum of node memory within the group.
{
Type: qbtypes.QueryTypeBuilder,
Spec: qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation]{
Name: "C",
Signal: telemetrytypes.SignalMetrics,
Aggregations: []qbtypes.MetricAggregation{
{
MetricName: "k8s.node.memory.working_set",
TimeAggregation: metrictypes.TimeAggregationAvg,
SpaceAggregation: metrictypes.SpaceAggregationSum,
ReduceTo: qbtypes.ReduceToAvg,
},
},
GroupBy: []qbtypes.GroupByKey{clusterNameGroupByKey},
Disabled: false,
},
},
// Query D: Memory allocatable — sum of node allocatable memory within the group.
// Same Latest caveat as Query B.
{
Type: qbtypes.QueryTypeBuilder,
Spec: qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation]{
Name: "D",
Signal: telemetrytypes.SignalMetrics,
Aggregations: []qbtypes.MetricAggregation{
{
MetricName: "k8s.node.allocatable_memory",
TimeAggregation: metrictypes.TimeAggregationLatest,
SpaceAggregation: metrictypes.SpaceAggregationSum,
ReduceTo: qbtypes.ReduceToAvg,
},
},
GroupBy: []qbtypes.GroupByKey{clusterNameGroupByKey},
Disabled: false,
},
},
}
return &qbtypes.QueryRangeRequest{
RequestType: qbtypes.RequestTypeScalar,
CompositeQuery: qbtypes.CompositeQuery{
Queries: queries,
},
}
}