feat: add bodySizeLimit to service and pod monitors (#6349)

* feat: add EnforcedBodySizeLimit to service and monitor

Documentation/api.md

@@ -1041,6 +1041,22 @@ string
 <p>The scrape class to apply.</p>
 </td>
 </tr>
+<tr>
+<td>
+<code>bodySizeLimit</code><br/>
+<em>
+<a href="#monitoring.coreos.com/v1.ByteSize">
+ByteSize
+</a>
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>When defined, bodySizeLimit specifies a job level limit on the size
+of uncompressed response body that will be accepted by Prometheus.</p>
+<p>It requires Prometheus &gt;= v2.28.0.</p>
+</td>
+</tr>
 </table>
 </td>
 </tr>
@@ -3384,6 +3400,22 @@ string
 <p>The scrape class to apply.</p>
 </td>
 </tr>
+<tr>
+<td>
+<code>bodySizeLimit</code><br/>
+<em>
+<a href="#monitoring.coreos.com/v1.ByteSize">
+ByteSize
+</a>
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>When defined, bodySizeLimit specifies a job level limit on the size
+of uncompressed response body that will be accepted by Prometheus.</p>
+<p>It requires Prometheus &gt;= v2.28.0.</p>
+</td>
+</tr>
 </table>
 </td>
 </tr>
@@ -5893,7 +5925,7 @@ authentication.</p>
 <h3 id="monitoring.coreos.com/v1.ByteSize">ByteSize
 (<code>string</code> alias)</h3>
 <p>
-(<em>Appears on:</em><a href="#monitoring.coreos.com/v1.CommonPrometheusFields">CommonPrometheusFields</a>, <a href="#monitoring.coreos.com/v1.PrometheusSpec">PrometheusSpec</a>)
+(<em>Appears on:</em><a href="#monitoring.coreos.com/v1.CommonPrometheusFields">CommonPrometheusFields</a>, <a href="#monitoring.coreos.com/v1.PodMonitorSpec">PodMonitorSpec</a>, <a href="#monitoring.coreos.com/v1.PrometheusSpec">PrometheusSpec</a>, <a href="#monitoring.coreos.com/v1.ServiceMonitorSpec">ServiceMonitorSpec</a>)
 </p>
 <div>
 <p>ByteSize is a valid memory size type based on powers-of-2, so 1KB is 1024B.
@@ -9438,6 +9470,22 @@ string
 <p>The scrape class to apply.</p>
 </td>
 </tr>
+<tr>
+<td>
+<code>bodySizeLimit</code><br/>
+<em>
+<a href="#monitoring.coreos.com/v1.ByteSize">
+ByteSize
+</a>
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>When defined, bodySizeLimit specifies a job level limit on the size
+of uncompressed response body that will be accepted by Prometheus.</p>
+<p>It requires Prometheus &gt;= v2.28.0.</p>
+</td>
+</tr>
 </tbody>
 </table>
 <h3 id="monitoring.coreos.com/v1.ProbeSpec">ProbeSpec
@@ -13717,6 +13765,22 @@ string
 <p>The scrape class to apply.</p>
 </td>
 </tr>
+<tr>
+<td>
+<code>bodySizeLimit</code><br/>
+<em>
+<a href="#monitoring.coreos.com/v1.ByteSize">
+ByteSize
+</a>
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>When defined, bodySizeLimit specifies a job level limit on the size
+of uncompressed response body that will be accepted by Prometheus.</p>
+<p>It requires Prometheus &gt;= v2.28.0.</p>
+</td>
+</tr>
 </tbody>
 </table>
 <h3 id="monitoring.coreos.com/v1.ShardStatus">ShardStatus

bundle.yaml

@@ -13532,6 +13532,12 @@ spec:
                       permission on the `Nodes` objects.
                     type: boolean
                 type: object
+              bodySizeLimit:
+                description: "When defined, bodySizeLimit specifies a job level limit
+                  on the size of uncompressed response body that will be accepted
+                  by Prometheus. \n It requires Prometheus >= v2.28.0."
+                pattern: (^0|([0-9]*[.])?[0-9]+((K|M|G|T|E|P)i?)?B)$
+                type: string
               jobLabel:
                 description: "The label to use to retrieve the job name from. `jobLabel`
                   selects the label from the associated Kubernetes `Pod` object which
@@ -37148,6 +37154,12 @@ spec:
                       permission on the `Nodes` objects.
                     type: boolean
                 type: object
+              bodySizeLimit:
+                description: "When defined, bodySizeLimit specifies a job level limit
+                  on the size of uncompressed response body that will be accepted
+                  by Prometheus. \n It requires Prometheus >= v2.28.0."
+                pattern: (^0|([0-9]*[.])?[0-9]+((K|M|G|T|E|P)i?)?B)$
+                type: string
               endpoints:
                 description: List of endpoints part of this ServiceMonitor.
                 items:

example/prometheus-operator-crd-full/monitoring.coreos.com_podmonitors.yaml

@@ -48,6 +48,12 @@ spec:
                       permission on the `Nodes` objects.
                     type: boolean
                 type: object
+              bodySizeLimit:
+                description: "When defined, bodySizeLimit specifies a job level limit
+                  on the size of uncompressed response body that will be accepted
+                  by Prometheus. \n It requires Prometheus >= v2.28.0."
+                pattern: (^0|([0-9]*[.])?[0-9]+((K|M|G|T|E|P)i?)?B)$
+                type: string
               jobLabel:
                 description: "The label to use to retrieve the job name from. `jobLabel`
                   selects the label from the associated Kubernetes `Pod` object which

example/prometheus-operator-crd-full/monitoring.coreos.com_servicemonitors.yaml

@@ -48,6 +48,12 @@ spec:
                       permission on the `Nodes` objects.
                     type: boolean
                 type: object
+              bodySizeLimit:
+                description: "When defined, bodySizeLimit specifies a job level limit
+                  on the size of uncompressed response body that will be accepted
+                  by Prometheus. \n It requires Prometheus >= v2.28.0."
+                pattern: (^0|([0-9]*[.])?[0-9]+((K|M|G|T|E|P)i?)?B)$
+                type: string
               endpoints:
                 description: List of endpoints part of this ServiceMonitor.
                 items:

example/prometheus-operator-crd/monitoring.coreos.com_podmonitors.yaml

@@ -49,6 +49,12 @@ spec:
                       permission on the `Nodes` objects.
                     type: boolean
                 type: object
+              bodySizeLimit:
+                description: "When defined, bodySizeLimit specifies a job level limit
+                  on the size of uncompressed response body that will be accepted
+                  by Prometheus. \n It requires Prometheus >= v2.28.0."
+                pattern: (^0|([0-9]*[.])?[0-9]+((K|M|G|T|E|P)i?)?B)$
+                type: string
               jobLabel:
                 description: "The label to use to retrieve the job name from. `jobLabel`
                   selects the label from the associated Kubernetes `Pod` object which

example/prometheus-operator-crd/monitoring.coreos.com_servicemonitors.yaml

@@ -49,6 +49,12 @@ spec:
                       permission on the `Nodes` objects.
                     type: boolean
                 type: object
+              bodySizeLimit:
+                description: "When defined, bodySizeLimit specifies a job level limit
+                  on the size of uncompressed response body that will be accepted
+                  by Prometheus. \n It requires Prometheus >= v2.28.0."
+                pattern: (^0|([0-9]*[.])?[0-9]+((K|M|G|T|E|P)i?)?B)$
+                type: string
               endpoints:
                 description: List of endpoints part of this ServiceMonitor.
                 items:

go.mod

@@ -4,6 +4,7 @@ go 1.22
 
 require (
 	github.com/alecthomas/kingpin/v2 v2.4.0
+	github.com/alecthomas/units v0.0.0-20231202071711-9a357b53e9c9
 	github.com/asaskevich/govalidator v0.0.0-20230301143203-a9d515a09cc2
 	github.com/blang/semver/v4 v4.0.0
 	github.com/cespare/xxhash/v2 v2.2.0
@@ -44,7 +45,6 @@ require (
 )
 
 require (
-	github.com/alecthomas/units v0.0.0-20231202071711-9a357b53e9c9 // indirect
 	github.com/bboreham/go-loser v0.0.0-20230920113527-fcc2c21820a3 // indirect
 	github.com/coreos/go-systemd/v22 v22.5.0 // indirect
 	github.com/facette/natsort v0.0.0-20181210072756-2cd4dd1e2dcb // indirect

jsonnet/prometheus-operator/podmonitors-crd.json

@@ -54,6 +54,11 @@
                     },
                     "type": "object"
                   },
+                  "bodySizeLimit": {
+                    "description": "When defined, bodySizeLimit specifies a job level limit on the size of uncompressed response body that will be accepted by Prometheus. \n It requires Prometheus >= v2.28.0.",
+                    "pattern": "(^0|([0-9]*[.])?[0-9]+((K|M|G|T|E|P)i?)?B)$",
+                    "type": "string"
+                  },
                   "jobLabel": {
                     "description": "The label to use to retrieve the job name from. `jobLabel` selects the label from the associated Kubernetes `Pod` object which will be used as the `job` label for all metrics. \n For example if `jobLabel` is set to `foo` and the Kubernetes `Pod` object is labeled with `foo: bar`, then Prometheus adds the `job=\"bar\"` label to all ingested metrics. \n If the value of this field is empty, the `job` label of the metrics defaults to the namespace and name of the PodMonitor object (e.g. `<namespace>/<name>`).",
                     "type": "string"

jsonnet/prometheus-operator/servicemonitors-crd.json

@@ -54,6 +54,11 @@
                     },
                     "type": "object"
                   },
+                  "bodySizeLimit": {
+                    "description": "When defined, bodySizeLimit specifies a job level limit on the size of uncompressed response body that will be accepted by Prometheus. \n It requires Prometheus >= v2.28.0.",
+                    "pattern": "(^0|([0-9]*[.])?[0-9]+((K|M|G|T|E|P)i?)?B)$",
+                    "type": "string"
+                  },
                   "endpoints": {
                     "description": "List of endpoints part of this ServiceMonitor.",
                     "items": {

pkg/apis/monitoring/v1/podmonitor_types.go

@@ -136,6 +136,14 @@ type PodMonitorSpec struct {
 	// +optional
 	// +kubebuilder:validation:MinLength=1
 	ScrapeClassName *string `json:"scrapeClass,omitempty"`
+
+	// When defined, bodySizeLimit specifies a job level limit on the size
+	// of uncompressed response body that will be accepted by Prometheus.
+	//
+	// It requires Prometheus >= v2.28.0.
+	//
+	// +optional
+	BodySizeLimit *ByteSize `json:"bodySizeLimit,omitempty"`
 }
 
 // PodMonitorList is a list of PodMonitors.

pkg/apis/monitoring/v1/servicemonitor_types.go

@@ -141,6 +141,14 @@ type ServiceMonitorSpec struct {
 	// +optional
 	// +kubebuilder:validation:MinLength=1
 	ScrapeClassName *string `json:"scrapeClass,omitempty"`
+
+	// When defined, bodySizeLimit specifies a job level limit on the size
+	// of uncompressed response body that will be accepted by Prometheus.
+	//
+	// It requires Prometheus >= v2.28.0.
+	//
+	// +optional
+	BodySizeLimit *ByteSize `json:"bodySizeLimit,omitempty"`
 }
 
 // ServiceMonitorList is a list of ServiceMonitors.

pkg/apis/monitoring/v1/zz_generated.deepcopy.go

@@ -1651,6 +1651,11 @@ func (in *PodMonitorSpec) DeepCopyInto(out *PodMonitorSpec) {
 		*out = new(string)
 		**out = **in
 	}
+	if in.BodySizeLimit != nil {
+		in, out := &in.BodySizeLimit, &out.BodySizeLimit
+		*out = new(ByteSize)
+		**out = **in
+	}
 }
 
 // DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PodMonitorSpec.
@@ -2818,6 +2823,11 @@ func (in *ServiceMonitorSpec) DeepCopyInto(out *ServiceMonitorSpec) {
 		*out = new(string)
 		**out = **in
 	}
+	if in.BodySizeLimit != nil {
+		in, out := &in.BodySizeLimit, &out.BodySizeLimit
+		*out = new(ByteSize)
+		**out = **in
+	}
 }
 
 // DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ServiceMonitorSpec.

pkg/client/applyconfiguration/monitoring/v1/podmonitorspec.go

@@ -38,6 +38,7 @@ type PodMonitorSpecApplyConfiguration struct {
 	KeepDroppedTargets    *uint64                                `json:"keepDroppedTargets,omitempty"`
 	AttachMetadata        *AttachMetadataApplyConfiguration      `json:"attachMetadata,omitempty"`
 	ScrapeClassName       *string                                `json:"scrapeClass,omitempty"`
+	BodySizeLimit         *monitoringv1.ByteSize                 `json:"bodySizeLimit,omitempty"`
 }
 
 // PodMonitorSpecApplyConfiguration constructs an declarative configuration of the PodMonitorSpec type for use with
@@ -166,3 +167,11 @@ func (b *PodMonitorSpecApplyConfiguration) WithScrapeClassName(value string) *Po
 	b.ScrapeClassName = &value
 	return b
 }
+
+// WithBodySizeLimit sets the BodySizeLimit field in the declarative configuration to the given value
+// and returns the receiver, so that objects can be built by chaining "With" function invocations.
+// If called multiple times, the BodySizeLimit field is set to the value of the last call.
+func (b *PodMonitorSpecApplyConfiguration) WithBodySizeLimit(value monitoringv1.ByteSize) *PodMonitorSpecApplyConfiguration {
+	b.BodySizeLimit = &value
+	return b
+}

pkg/client/applyconfiguration/monitoring/v1/servicemonitorspec.go

@@ -39,6 +39,7 @@ type ServiceMonitorSpecApplyConfiguration struct {
 	KeepDroppedTargets    *uint64                              `json:"keepDroppedTargets,omitempty"`
 	AttachMetadata        *AttachMetadataApplyConfiguration    `json:"attachMetadata,omitempty"`
 	ScrapeClassName       *string                              `json:"scrapeClass,omitempty"`
+	BodySizeLimit         *monitoringv1.ByteSize               `json:"bodySizeLimit,omitempty"`
 }
 
 // ServiceMonitorSpecApplyConfiguration constructs an declarative configuration of the ServiceMonitorSpec type for use with
@@ -177,3 +178,11 @@ func (b *ServiceMonitorSpecApplyConfiguration) WithScrapeClassName(value string)
 	b.ScrapeClassName = &value
 	return b
 }
+
+// WithBodySizeLimit sets the BodySizeLimit field in the declarative configuration to the given value
+// and returns the receiver, so that objects can be built by chaining "With" function invocations.
+// If called multiple times, the BodySizeLimit field is set to the value of the last call.
+func (b *ServiceMonitorSpecApplyConfiguration) WithBodySizeLimit(value monitoringv1.ByteSize) *ServiceMonitorSpecApplyConfiguration {
+	b.BodySizeLimit = &value
+	return b
+}

pkg/prometheus/promcfg.go

@@ -22,6 +22,7 @@ import (
 	"sort"
 	"strings"
 
+	"github.com/alecthomas/units"
 	"github.com/blang/semver/v4"
 	"github.com/go-kit/log"
 	"github.com/go-kit/log/level"
@@ -1034,8 +1035,8 @@ func (cg *ConfigGenerator) generatePodMonitorConfig(
 	cfg = cg.AddLimitsToYAML(cfg, keepDroppedTargetsKey, m.Spec.KeepDroppedTargets, cpf.EnforcedKeepDroppedTargets)
 	cfg = cg.AddScrapeProtocols(cfg, m.Spec.ScrapeProtocols)
 
-	if cpf.EnforcedBodySizeLimit != "" {
-		cfg = cg.WithMinimumVersion("2.28.0").AppendMapItem(cfg, "body_size_limit", cpf.EnforcedBodySizeLimit)
+	if bodySizeLimit := getLowerByteSize(m.Spec.BodySizeLimit, &cpf); !isByteSizeEmpty(bodySizeLimit) {
+		cfg = cg.WithMinimumVersion("2.28.0").AppendMapItem(cfg, "body_size_limit", bodySizeLimit)
 	}
 
 	cfg = append(cfg, yaml.MapItem{Key: "metric_relabel_configs", Value: generateRelabelConfig(labeler.GetRelabelingConfigs(m.TypeMeta, m.ObjectMeta, ep.MetricRelabelConfigs))})
@@ -1546,8 +1547,8 @@ func (cg *ConfigGenerator) generateServiceMonitorConfig(
 	cfg = cg.AddLimitsToYAML(cfg, keepDroppedTargetsKey, m.Spec.KeepDroppedTargets, cpf.EnforcedKeepDroppedTargets)
 	cfg = cg.AddScrapeProtocols(cfg, m.Spec.ScrapeProtocols)
 
-	if cpf.EnforcedBodySizeLimit != "" {
-		cfg = cg.WithMinimumVersion("2.28.0").AppendMapItem(cfg, "body_size_limit", cpf.EnforcedBodySizeLimit)
+	if bodySizeLimit := getLowerByteSize(m.Spec.BodySizeLimit, &cpf); !isByteSizeEmpty(bodySizeLimit) {
+		cfg = cg.WithMinimumVersion("2.28.0").AppendMapItem(cfg, "body_size_limit", bodySizeLimit)
 	}
 
 	cfg = append(cfg, yaml.MapItem{Key: "metric_relabel_configs", Value: generateRelabelConfig(labeler.GetRelabelingConfigs(m.TypeMeta, m.ObjectMeta, ep.MetricRelabelConfigs))})
@@ -3413,3 +3414,26 @@ func (cg *ConfigGenerator) getScrapeClassOrDefault(name *string) *monitoringv1.S
 	}
 	return nil
 }
+
+func getLowerByteSize(v *monitoringv1.ByteSize, cpf *monitoringv1.CommonPrometheusFields) *monitoringv1.ByteSize {
+	if isByteSizeEmpty(&cpf.EnforcedBodySizeLimit) {
+		return v
+	}
+
+	if isByteSizeEmpty(v) {
+		return &cpf.EnforcedBodySizeLimit
+	}
+
+	vBytes, _ := units.ParseBase2Bytes(string(*v))
+	pBytes, _ := units.ParseBase2Bytes(string(cpf.EnforcedBodySizeLimit))
+
+	if vBytes > pBytes {
+		return &cpf.EnforcedBodySizeLimit
+	}
+
+	return v
+}
+
+func isByteSizeEmpty(v *monitoringv1.ByteSize) bool {
+	return v == nil || *v == ""
+}

pkg/prometheus/promcfg_test.go

@@ -4356,21 +4356,29 @@ func TestKeepDroppedTargets(t *testing.T) {
 
 func TestBodySizeLimits(t *testing.T) {
 	for _, tc := range []struct {
-		version               string
-		enforcedBodySizeLimit monitoringv1.ByteSize
-		expectedErr           error
-		golden                string
+		version                     string
+		enforcedBodySizeLimit       monitoringv1.ByteSize
+		serviceMonitorBodySizeLimit *monitoringv1.ByteSize
+		expectedErr                 error
+		golden                      string
 	}{
 		{
-			version:               "v2.27.0",
-			enforcedBodySizeLimit: "1000MB",
-			golden:                "BodySizeLimits_enforce1000MB_v2.27.0.golden",
+			version:                     "v2.27.0",
+			enforcedBodySizeLimit:       "1000MB",
+			serviceMonitorBodySizeLimit: ptr.To[monitoringv1.ByteSize]("2GB"),
+			golden:                      "BodySizeLimits_enforce_v2.27.0.golden",
 		},
 		{
 			version:               "v2.28.0",
 			enforcedBodySizeLimit: "1000MB",
 			golden:                "BodySizeLimits_enforce1000MB_v2.28.0.golden",
 		},
+		{
+			version:                     "v2.28.0",
+			enforcedBodySizeLimit:       "4000MB",
+			serviceMonitorBodySizeLimit: ptr.To[monitoringv1.ByteSize]("2GB"),
+			golden:                      "BodySizeLimits_enforce2GB_v2.28.0.golden",
+		},
 		{
 			version:               "v2.28.0",
 			enforcedBodySizeLimit: "",
@@ -4400,6 +4408,7 @@ func TestBodySizeLimits(t *testing.T) {
 							Interval: "30s",
 						},
 					},
+					BodySizeLimit: tc.serviceMonitorBodySizeLimit,
 				},
 			}
 

pkg/prometheus/testdata/BodySizeLimits_enforce2GB_v2.28.0.golden

@@ -0,0 +1,70 @@
+global:
+  evaluation_interval: 30s
+  scrape_interval: 30s
+  external_labels:
+    prometheus: default/test
+    prometheus_replica: $(POD_NAME)
+scrape_configs:
+- job_name: serviceMonitor/default/testservicemonitor1/0
+  honor_labels: false
+  kubernetes_sd_configs:
+  - role: endpoints
+    namespaces:
+      names:
+      - default
+  scrape_interval: 30s
+  relabel_configs:
+  - source_labels:
+    - job
+    target_label: __tmp_prometheus_job_name
+  - action: keep
+    source_labels:
+    - __meta_kubernetes_endpoint_port_name
+    regex: web
+  - source_labels:
+    - __meta_kubernetes_endpoint_address_target_kind
+    - __meta_kubernetes_endpoint_address_target_name
+    separator: ;
+    regex: Node;(.*)
+    replacement: ${1}
+    target_label: node
+  - source_labels:
+    - __meta_kubernetes_endpoint_address_target_kind
+    - __meta_kubernetes_endpoint_address_target_name
+    separator: ;
+    regex: Pod;(.*)
+    replacement: ${1}
+    target_label: pod
+  - source_labels:
+    - __meta_kubernetes_namespace
+    target_label: namespace
+  - source_labels:
+    - __meta_kubernetes_service_name
+    target_label: service
+  - source_labels:
+    - __meta_kubernetes_pod_name
+    target_label: pod
+  - source_labels:
+    - __meta_kubernetes_pod_container_name
+    target_label: container
+  - action: drop
+    source_labels:
+    - __meta_kubernetes_pod_phase
+    regex: (Failed|Succeeded)
+  - source_labels:
+    - __meta_kubernetes_service_name
+    target_label: job
+    replacement: ${1}
+  - target_label: endpoint
+    replacement: web
+  - source_labels:
+    - __address__
+    target_label: __tmp_hash
+    modulus: 1
+    action: hashmod
+  - source_labels:
+    - __tmp_hash
+    regex: $(SHARD)
+    action: keep
+  body_size_limit: 2GB
+  metric_relabel_configs: []