Merge pull request #401 from zalando-incubator/update-deps-02-14

Update dependencies
Use cdp-runtime for build (#398 )
2026-05-21 02:44:39 +00:00 · 2022-02-14 15:05:19 +01:00 · 2022-02-14 14:41:57 +01:00 · 2022-02-14 14:36:06 +01:00 · 2022-02-01 18:13:48 +01:00 · 2022-01-12 11:36:53 +01:00
24 changed files with 2199 additions and 622 deletions
@@ -1,4 +1,4 @@
-FROM registry.opensource.zalan.do/library/alpine-3.12:latest
+FROM registry.opensource.zalan.do/library/alpine-3.13:latest
 LABEL maintainer="Team Teapot @ Zalando SE <team-teapot@zalando.de>"

 RUN apk add --no-cache tzdata
@@ -720,12 +720,63 @@ The `ScalingSchedule` and `ClusterScalingSchedule` collectors allow
 collecting time-based metrics from the respective CRD objects specified
 in the HPA.

+These collectors are disabled by default, you have to start the server
+with the `--scaling-schedule` flag to enable it. Remember to deploy the CRDs
+`ScalingSchedule` and `ClusterScalingSchedule` and allow the service
+account used by the server to read, watch and list them.
+
 ### Supported metrics

 | Metric | Description | Type | K8s Versions |
 | ---------- | -------------- | ------- | -- |
 | ObjectName | The metric is calculated and stored for each `ScalingSchedule` and `ClusterScalingSchedule` referenced in the HPAs | `ScalingSchedule` and `ClusterScalingSchedule` | `>=1.16` |

+### Ramp-up and ramp-down feature
+
+To avoid abrupt scaling due to time based metrics,the `SchalingSchedule`
+collector has a feature of ramp-up and ramp-down the metric over a
+specific period of time. The duration of the scaling window can be
+configured individually in the `[Cluster]ScalingSchedule` object, via
+the option `scalingWindowDurationMinutes` or globally for all scheduled
+events, and defaults to a globally configured value if not specified.
+The default for the latter is set to 10 minutes, but can be changed
+using the `--scaling-schedule-default-scaling-window` flag.
+
+This spreads the scale events around, creating less load on the other
+components, and helping the rest of the metrics (like the CPU ones) to
+adjust as well.
+
+The [HPA algorithm][algo-details] does not make changes if the metric
+change is less than the specified by the
+`horizontal-pod-autoscaler-tolerance` flag:
+
+> We'll skip scaling if the ratio is sufficiently close to 1.0 (within a
+> globally-configurable tolerance, from the
+> `--horizontal-pod-autoscaler-tolerance` flag, which defaults to 0.1.
+
+With that in mind, the ramp-up and ramp-down feature divides the scaling
+over the specified period of time in buckets, trying to achieve changes
+bigger than the configured tolerance. The number of buckets defaults to
+10 and can be configured by the `--scaling-schedule-ramp-steps` flag.
+
+**Important**: note that the ramp-up and ramp-down feature can lead to
+deployments achieving less than the specified number of pods, due to the
+HPA 10% change rule and the ceiling function applied to the desired
+number of the pods (check the [algorithm details][algo-details]). It
+varies with the configured metric for `ScalingSchedule` events, the
+number of pods and the configured `horizontal-pod-autoscaler-tolerance`
+flag of your kubernetes installation. [This gist][gist] contains the code to
+simulate the situations a deployment with different number of pods, with
+a metric of 10000 can face with 10 buckets (max of 90% of the metric
+returned) and 5 buckets (max of 80% of the metric returned). The ramp-up
+and ramp-down feature can be disabled by setting
+`--scaling-schedule-default-scaling-window` to 0 and abrupt scalings can
+be handled via [scaling policies][policies].
+
+[algo-details]: https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/#algorithm-details
+[gist]: https://gist.github.com/jonathanbeber/37f1f918ab7ef6101c6ce56cc2cef3a2
+[policies]: https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/#scaling-policies
+
 ### Example

 This is an example of using the ScalingSchedule collectors to collect
@@ -826,8 +877,3 @@ Note that these number of pods are just considering these custom
 metrics, the normal HPA behavior still applies, such as: in case of
 multiple metrics the biggest number of pods is the utilized one, HPA max
 and min replica configuration, autoscaling policies, etc.
-
-These collectors are disabled by default, you have to start the server
-with the `--scaling-schedule` flag to enable it. Remember to deploy the CRDs
-`ScalingSchedule` and `ClusterScalingSchedule` and allow the service
-account used by the server to read, watch and list them.
@@ -1,7 +1,9 @@
 version: "2017-09-20"
 pipeline:
 - id: build
-  overlay: ci/golang
+  vm_config:
+    type: linux
+    image: "cdp-runtime/go"
  cache:
    paths:
    - /go/pkg/mod       # pkg cache for Go modules
@@ -1,10 +1,9 @@
-
 ---
 apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
  annotations:
-    controller-gen.kubebuilder.io/version: v0.5.0
+    controller-gen.kubebuilder.io/version: v0.8.0
  creationTimestamp: null
  name: clusterscalingschedules.zalando.org
 spec:
@@ -1,10 +1,9 @@
-
 ---
 apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
  annotations:
-    controller-gen.kubebuilder.io/version: v0.5.0
+    controller-gen.kubebuilder.io/version: v0.8.0
  creationTimestamp: null
  name: scalingschedules.zalando.org
 spec:
@@ -1,5 +1,5 @@
-FROM registry.opensource.zalan.do/stups/alpine:latest
-MAINTAINER Team Teapot @ Zalando SE <team-teapot@zalando.de>
+FROM registry.opensource.zalan.do/library/alpine-3.13:latest
+LABEL maintainer="Team Teapot @ Zalando SE <team-teapot@zalando.de>"

 # add binary
 ADD build/linux/custom-metrics-consumer /
@@ -11,7 +11,9 @@ import (
 func metricsHandler(w http.ResponseWriter, r *http.Request) {
 	w.WriteHeader(200)
 	_, err := w.Write([]byte(fmt.Sprintf(`{"queue": {"length": %d}}`, size)))
-	log.Fatalf("failed to write: %v", err)
+	if err != nil {
+		log.Fatalf("failed to write: %v", err)
+	}
 }

 var (
@@ -1,35 +1,123 @@
 module github.com/zalando-incubator/kube-metrics-adapter

 require (
-	github.com/NYTimes/gziphandler v1.0.1 // indirect
-	github.com/aws/aws-sdk-go v1.40.22
-	github.com/go-openapi/spec v0.20.3
+	github.com/aws/aws-sdk-go v1.42.52
 	github.com/influxdata/influxdb-client-go v0.2.0
-	github.com/influxdata/line-protocol v0.0.0-20201012155213-5f565037cbc9 // indirect
 	github.com/kubernetes-sigs/custom-metrics-apiserver v0.0.0-20201216091021-1b9fa998bbaa
-	github.com/mailru/easyjson v0.7.7 // indirect
-	github.com/prometheus/client_golang v1.11.0
-	github.com/prometheus/common v0.26.0
+	github.com/prometheus/client_golang v1.12.1
+	github.com/prometheus/common v0.32.1
 	github.com/sirupsen/logrus v1.8.1
-	github.com/spf13/cobra v1.1.3
-	github.com/spyzhov/ajson v0.4.2
+	github.com/spf13/cobra v1.3.0
+	github.com/spyzhov/ajson v0.7.1
 	github.com/stretchr/testify v1.7.0
-	github.com/szuecs/routegroup-client v0.18.3
-	github.com/zalando-incubator/cluster-lifecycle-manager v0.0.0-20180921141935-824b77fb1f84
-	go.uber.org/zap v1.13.0 // indirect
-	golang.org/x/crypto v0.0.0-20201124201722-c8d3bf9c5392 // indirect
-	golang.org/x/oauth2 v0.0.0-20200107190931-bf48bf16ab8d
-	gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b // indirect
-	k8s.io/api v0.20.9
-	k8s.io/apimachinery v0.20.9
-	k8s.io/apiserver v0.20.9
-	k8s.io/client-go v0.20.9
-	k8s.io/code-generator v0.20.9
-	k8s.io/component-base v0.20.9
+	github.com/szuecs/routegroup-client v0.21.0
+	github.com/zalando-incubator/cluster-lifecycle-manager v0.0.0-20220201095549-bbdeecaa4fc1
+	golang.org/x/oauth2 v0.0.0-20211104180415-d3ed0bb246c8
+	k8s.io/api v0.23.0
+	k8s.io/apimachinery v0.23.0
+	k8s.io/apiserver v0.23.0
+	k8s.io/client-go v0.23.0
+	k8s.io/code-generator v0.23.0
+	k8s.io/component-base v0.23.0
 	k8s.io/klog v1.0.0
-	k8s.io/kube-openapi v0.0.0-20201113171705-d219536bb9fd
-	k8s.io/metrics v0.20.9
-	sigs.k8s.io/controller-tools v0.5.0
+	k8s.io/kube-openapi v0.0.0-20220124234850-424119656bbf
+	k8s.io/metrics v0.21.9
+	sigs.k8s.io/controller-tools v0.8.0
 )

-go 1.16
+require (
+	github.com/NYTimes/gziphandler v1.1.1 // indirect
+	github.com/PuerkitoBio/purell v1.1.1 // indirect
+	github.com/PuerkitoBio/urlesc v0.0.0-20170810143723-de5bf2ad4578 // indirect
+	github.com/beorn7/perks v1.0.1 // indirect
+	github.com/blang/semver v3.5.1+incompatible // indirect
+	github.com/cespare/xxhash/v2 v2.1.2 // indirect
+	github.com/coreos/go-semver v0.3.0 // indirect
+	github.com/coreos/go-systemd/v22 v22.3.2 // indirect
+	github.com/davecgh/go-spew v1.1.1 // indirect
+	github.com/emicklei/go-restful v2.14.3+incompatible // indirect
+	github.com/evanphx/json-patch v4.12.0+incompatible // indirect
+	github.com/fatih/color v1.13.0 // indirect
+	github.com/felixge/httpsnoop v1.0.2 // indirect
+	github.com/fsnotify/fsnotify v1.5.1 // indirect
+	github.com/go-logr/logr v1.2.2 // indirect
+	github.com/go-openapi/jsonpointer v0.19.5 // indirect
+	github.com/go-openapi/jsonreference v0.19.6 // indirect
+	github.com/go-openapi/swag v0.21.1 // indirect
+	github.com/gobuffalo/flect v0.2.3 // indirect
+	github.com/gogo/protobuf v1.3.2 // indirect
+	github.com/golang-jwt/jwt v3.2.2+incompatible // indirect
+	github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da // indirect
+	github.com/golang/protobuf v1.5.2 // indirect
+	github.com/google/go-cmp v0.5.7 // indirect
+	github.com/google/gofuzz v1.2.0 // indirect
+	github.com/google/uuid v1.3.0 // indirect
+	github.com/googleapis/gnostic v0.5.5 // indirect
+	github.com/grpc-ecosystem/go-grpc-prometheus v1.2.0 // indirect
+	github.com/grpc-ecosystem/grpc-gateway v1.16.0 // indirect
+	github.com/imdario/mergo v0.3.6 // indirect
+	github.com/inconshreveable/mousetrap v1.0.0 // indirect
+	github.com/influxdata/line-protocol v0.0.0-20210922203350-b1ad95c89adf // indirect
+	github.com/jmespath/go-jmespath v0.4.0 // indirect
+	github.com/josharian/intern v1.0.0 // indirect
+	github.com/json-iterator/go v1.1.12 // indirect
+	github.com/mailru/easyjson v0.7.7 // indirect
+	github.com/mattn/go-colorable v0.1.12 // indirect
+	github.com/mattn/go-isatty v0.0.14 // indirect
+	github.com/matttproud/golang_protobuf_extensions v1.0.2-0.20181231171920-c182affec369 // indirect
+	github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
+	github.com/modern-go/reflect2 v1.0.2 // indirect
+	github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 // indirect
+	github.com/pkg/errors v0.9.1 // indirect
+	github.com/pmezard/go-difflib v1.0.0 // indirect
+	github.com/prometheus/client_model v0.2.0 // indirect
+	github.com/prometheus/procfs v0.7.3 // indirect
+	github.com/spf13/afero v1.8.1 // indirect
+	github.com/spf13/pflag v1.0.5 // indirect
+	go.etcd.io/etcd/api/v3 v3.5.1 // indirect
+	go.etcd.io/etcd/client/pkg/v3 v3.5.1 // indirect
+	go.etcd.io/etcd/client/v3 v3.5.0 // indirect
+	go.opentelemetry.io/contrib v0.20.0 // indirect
+	go.opentelemetry.io/contrib/instrumentation/google.golang.org/grpc/otelgrpc v0.20.0 // indirect
+	go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.20.0 // indirect
+	go.opentelemetry.io/otel v0.20.0 // indirect
+	go.opentelemetry.io/otel/exporters/otlp v0.20.0 // indirect
+	go.opentelemetry.io/otel/metric v0.20.0 // indirect
+	go.opentelemetry.io/otel/sdk v0.20.0 // indirect
+	go.opentelemetry.io/otel/sdk/export/metric v0.20.0 // indirect
+	go.opentelemetry.io/otel/sdk/metric v0.20.0 // indirect
+	go.opentelemetry.io/otel/trace v0.20.0 // indirect
+	go.opentelemetry.io/proto/otlp v0.7.0 // indirect
+	go.uber.org/atomic v1.7.0 // indirect
+	go.uber.org/multierr v1.6.0 // indirect
+	go.uber.org/zap v1.19.0 // indirect
+	golang.org/x/crypto v0.0.0-20220131195533-30dcbda58838 // indirect
+	golang.org/x/mod v0.5.1 // indirect
+	golang.org/x/net v0.0.0-20220127200216-cd36cc0744dd // indirect
+	golang.org/x/sync v0.0.0-20210220032951-036812b2e83c // indirect
+	golang.org/x/sys v0.0.0-20220209214540-3681064d5158 // indirect
+	golang.org/x/term v0.0.0-20210927222741-03fcf44c2211 // indirect
+	golang.org/x/text v0.3.7 // indirect
+	golang.org/x/time v0.0.0-20220210224613-90d013bbcef8 // indirect
+	golang.org/x/tools v0.1.9 // indirect
+	golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 // indirect
+	google.golang.org/appengine v1.6.7 // indirect
+	google.golang.org/genproto v0.0.0-20211208223120-3a66f561d7aa // indirect
+	google.golang.org/grpc v1.43.0 // indirect
+	google.golang.org/protobuf v1.27.1 // indirect
+	gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c // indirect
+	gopkg.in/inf.v0 v0.9.1 // indirect
+	gopkg.in/natefinch/lumberjack.v2 v2.0.0 // indirect
+	gopkg.in/yaml.v2 v2.4.0 // indirect
+	gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b // indirect
+	k8s.io/apiextensions-apiserver v0.23.0 // indirect
+	k8s.io/gengo v0.0.0-20210813121822-485abfe95c7c // indirect
+	k8s.io/klog/v2 v2.40.1 // indirect
+	k8s.io/utils v0.0.0-20220210201930-3a6ce19ff2f9 // indirect
+	sigs.k8s.io/apiserver-network-proxy/konnectivity-client v0.0.27 // indirect
+	sigs.k8s.io/json v0.0.0-20211208200746-9f7c6b3444d2 // indirect
+	sigs.k8s.io/structured-merge-diff/v4 v4.2.1 // indirect
+	sigs.k8s.io/yaml v1.3.0 // indirect
+)
+
+go 1.17
@@ -1,3 +1,4 @@
+//go:build !ignore_autogenerated
 // +build !ignore_autogenerated

 /*
@@ -20,6 +20,7 @@ package versioned

 import (
 	"fmt"
+	"net/http"

 	zalandov1 "github.com/zalando-incubator/kube-metrics-adapter/pkg/client/clientset/versioned/typed/zalando.org/v1"
 	discovery "k8s.io/client-go/discovery"
@@ -55,22 +56,41 @@ func (c *Clientset) Discovery() discovery.DiscoveryInterface {
 // NewForConfig creates a new Clientset for the given config.
 // If config's RateLimiter is not set and QPS and Burst are acceptable,
 // NewForConfig will generate a rate-limiter in configShallowCopy.
+// NewForConfig is equivalent to NewForConfigAndClient(c, httpClient),
+// where httpClient was generated with rest.HTTPClientFor(c).
 func NewForConfig(c *rest.Config) (*Clientset, error) {
 	configShallowCopy := *c
+
+	// share the transport between all clients
+	httpClient, err := rest.HTTPClientFor(&configShallowCopy)
+	if err != nil {
+		return nil, err
+	}
+
+	return NewForConfigAndClient(&configShallowCopy, httpClient)
+}
+
+// NewForConfigAndClient creates a new Clientset for the given config and http client.
+// Note the http client provided takes precedence over the configured transport values.
+// If config's RateLimiter is not set and QPS and Burst are acceptable,
+// NewForConfigAndClient will generate a rate-limiter in configShallowCopy.
+func NewForConfigAndClient(c *rest.Config, httpClient *http.Client) (*Clientset, error) {
+	configShallowCopy := *c
 	if configShallowCopy.RateLimiter == nil && configShallowCopy.QPS > 0 {
 		if configShallowCopy.Burst <= 0 {
 			return nil, fmt.Errorf("burst is required to be greater than 0 when RateLimiter is not set and QPS is set to greater than 0")
 		}
 		configShallowCopy.RateLimiter = flowcontrol.NewTokenBucketRateLimiter(configShallowCopy.QPS, configShallowCopy.Burst)
 	}
+
 	var cs Clientset
 	var err error
-	cs.zalandoV1, err = zalandov1.NewForConfig(&configShallowCopy)
+	cs.zalandoV1, err = zalandov1.NewForConfigAndClient(&configShallowCopy, httpClient)
 	if err != nil {
 		return nil, err
 	}

-	cs.DiscoveryClient, err = discovery.NewDiscoveryClientForConfig(&configShallowCopy)
+	cs.DiscoveryClient, err = discovery.NewDiscoveryClientForConfigAndClient(&configShallowCopy, httpClient)
 	if err != nil {
 		return nil, err
 	}
@@ -80,11 +100,11 @@ func NewForConfig(c *rest.Config) (*Clientset, error) {
 // NewForConfigOrDie creates a new Clientset for the given config and
 // panics if there is an error in the config.
 func NewForConfigOrDie(c *rest.Config) *Clientset {
-	var cs Clientset
-	cs.zalandoV1 = zalandov1.NewForConfigOrDie(c)
-
-	cs.DiscoveryClient = discovery.NewDiscoveryClientForConfigOrDie(c)
-	return &cs
+	cs, err := NewForConfig(c)
+	if err != nil {
+		panic(err)
+	}
+	return cs
 }

 // New creates a new Clientset for the given RESTClient.
@@ -74,7 +74,10 @@ func (c *Clientset) Tracker() testing.ObjectTracker {
 	return c.tracker
 }

-var _ clientset.Interface = &Clientset{}
+var (
+	_ clientset.Interface = &Clientset{}
+	_ testing.FakeClient  = &Clientset{}
+)

 // ZalandoV1 retrieves the ZalandoV1Client
 func (c *Clientset) ZalandoV1() zalandov1.ZalandoV1Interface {
@@ -99,7 +99,7 @@ func (c *FakeClusterScalingSchedules) Update(ctx context.Context, clusterScaling
 // Delete takes name of the clusterScalingSchedule and deletes it. Returns an error if one occurs.
 func (c *FakeClusterScalingSchedules) Delete(ctx context.Context, name string, opts v1.DeleteOptions) error {
 	_, err := c.Fake.
-		Invokes(testing.NewRootDeleteAction(clusterscalingschedulesResource, name), &zalandoorgv1.ClusterScalingSchedule{})
+		Invokes(testing.NewRootDeleteActionWithOptions(clusterscalingschedulesResource, name, opts), &zalandoorgv1.ClusterScalingSchedule{})
 	return err
 }

@@ -105,7 +105,7 @@ func (c *FakeScalingSchedules) Update(ctx context.Context, scalingSchedule *zala
 // Delete takes name of the scalingSchedule and deletes it. Returns an error if one occurs.
 func (c *FakeScalingSchedules) Delete(ctx context.Context, name string, opts v1.DeleteOptions) error {
 	_, err := c.Fake.
-		Invokes(testing.NewDeleteAction(scalingschedulesResource, c.ns, name), &zalandoorgv1.ScalingSchedule{})
+		Invokes(testing.NewDeleteActionWithOptions(scalingschedulesResource, c.ns, name, opts), &zalandoorgv1.ScalingSchedule{})

 	return err
 }
@@ -19,6 +19,8 @@ limitations under the License.
 package v1

 import (
+	"net/http"
+
 	v1 "github.com/zalando-incubator/kube-metrics-adapter/pkg/apis/zalando.org/v1"
 	"github.com/zalando-incubator/kube-metrics-adapter/pkg/client/clientset/versioned/scheme"
 	rest "k8s.io/client-go/rest"
@@ -44,12 +46,28 @@ func (c *ZalandoV1Client) ScalingSchedules(namespace string) ScalingScheduleInte
 }

 // NewForConfig creates a new ZalandoV1Client for the given config.
+// NewForConfig is equivalent to NewForConfigAndClient(c, httpClient),
+// where httpClient was generated with rest.HTTPClientFor(c).
 func NewForConfig(c *rest.Config) (*ZalandoV1Client, error) {
 	config := *c
 	if err := setConfigDefaults(&config); err != nil {
 		return nil, err
 	}
-	client, err := rest.RESTClientFor(&config)
+	httpClient, err := rest.HTTPClientFor(&config)
+	if err != nil {
+		return nil, err
+	}
+	return NewForConfigAndClient(&config, httpClient)
+}
+
+// NewForConfigAndClient creates a new ZalandoV1Client for the given config and http client.
+// Note the http client provided takes precedence over the configured transport values.
+func NewForConfigAndClient(c *rest.Config, h *http.Client) (*ZalandoV1Client, error) {
+	config := *c
+	if err := setConfigDefaults(&config); err != nil {
+		return nil, err
+	}
+	client, err := rest.RESTClientForConfigAndClient(&config, h)
 	if err != nil {
 		return nil, err
 	}
@@ -72,10 +72,25 @@ func (g *JSONPathMetricsGetter) GetMetric(metricsURL url.URL) (float64, error) {
 		return 0, err
 	}

-	if len(nodes) != 1 {
+	if len(nodes) == 0 {
 		return 0, fmt.Errorf("unexpected json: expected single numeric or array value")
 	}

+	if len(nodes) > 1 {
+		if g.aggregator == nil {
+			return 0, fmt.Errorf("no aggregator function has been specified")
+		}
+		values := make([]float64, 0, len(nodes))
+		for _, node := range nodes {
+			v, err := node.GetNumeric()
+			if err != nil {
+				return 0, fmt.Errorf("unexpected json: did not find numeric or array value '%s': %w", nodes, err)
+			}
+			values = append(values, v)
+		}
+		return g.aggregator(values...), nil
+	}
+
 	node := nodes[0]
 	if node.IsArray() {
 		if g.aggregator == nil {
@@ -51,6 +51,13 @@ func TestJSONPathMetricsGetter(t *testing.T) {
 			result:       5,
 			aggregator:   Average,
 		},
+		{
+			name:         "glob array query",
+			jsonResponse: []byte(`{"worker_status":[{"last_status":{"backlog":3}},{"last_status":{"backlog":7}}]}`),
+			jsonPath:     "$.worker_status.[*].last_status.backlog",
+			result:       5,
+			aggregator:   Average,
+		},
 		{
 			name:         "json path not resulting in array or number should lead to error",
 			jsonResponse: []byte(`{"metric.value":5}`),
@@ -82,6 +82,7 @@ type ScalingScheduleCollectorPlugin struct {
 	store                Store
 	now                  Now
 	defaultScalingWindow time.Duration
+	rampSteps            int
 }

 // ClusterScalingScheduleCollectorPlugin is a collector plugin for initializing metrics
@@ -90,23 +91,26 @@ type ClusterScalingScheduleCollectorPlugin struct {
 	store                Store
 	now                  Now
 	defaultScalingWindow time.Duration
+	rampSteps            int
 }

 // NewScalingScheduleCollectorPlugin initializes a new ScalingScheduleCollectorPlugin.
-func NewScalingScheduleCollectorPlugin(store Store, now Now, defaultScalingWindow time.Duration) (*ScalingScheduleCollectorPlugin, error) {
+func NewScalingScheduleCollectorPlugin(store Store, now Now, defaultScalingWindow time.Duration, rampSteps int) (*ScalingScheduleCollectorPlugin, error) {
 	return &ScalingScheduleCollectorPlugin{
 		store:                store,
 		now:                  now,
 		defaultScalingWindow: defaultScalingWindow,
+		rampSteps:            rampSteps,
 	}, nil
 }

 // NewClusterScalingScheduleCollectorPlugin initializes a new ClusterScalingScheduleCollectorPlugin.
-func NewClusterScalingScheduleCollectorPlugin(store Store, now Now, defaultScalingWindow time.Duration) (*ClusterScalingScheduleCollectorPlugin, error) {
+func NewClusterScalingScheduleCollectorPlugin(store Store, now Now, defaultScalingWindow time.Duration, rampSteps int) (*ClusterScalingScheduleCollectorPlugin, error) {
 	return &ClusterScalingScheduleCollectorPlugin{
 		store:                store,
 		now:                  now,
 		defaultScalingWindow: defaultScalingWindow,
+		rampSteps:            rampSteps,
 	}, nil
 }

@@ -114,14 +118,14 @@ func NewClusterScalingScheduleCollectorPlugin(store Store, now Now, defaultScali
 // specified HPA. It's the only required method to implement the
 // collector.CollectorPlugin interface.
 func (c *ScalingScheduleCollectorPlugin) NewCollector(hpa *autoscalingv2.HorizontalPodAutoscaler, config *MetricConfig, interval time.Duration) (Collector, error) {
-	return NewScalingScheduleCollector(c.store, c.defaultScalingWindow, c.now, hpa, config, interval)
+	return NewScalingScheduleCollector(c.store, c.defaultScalingWindow, c.rampSteps, c.now, hpa, config, interval)
 }

 // NewCollector initializes a new cluster wide scaling schedule
 // collector from the specified HPA. It's the only required method to
 // implement the collector.CollectorPlugin interface.
 func (c *ClusterScalingScheduleCollectorPlugin) NewCollector(hpa *autoscalingv2.HorizontalPodAutoscaler, config *MetricConfig, interval time.Duration) (Collector, error) {
-	return NewClusterScalingScheduleCollector(c.store, c.defaultScalingWindow, c.now, hpa, config, interval)
+	return NewClusterScalingScheduleCollector(c.store, c.defaultScalingWindow, c.rampSteps, c.now, hpa, config, interval)
 }

 // ScalingScheduleCollector is a metrics collector for time based
@@ -148,10 +152,11 @@ type scalingScheduleCollector struct {
 	interval             time.Duration
 	config               MetricConfig
 	defaultScalingWindow time.Duration
+	rampSteps            int
 }

 // NewScalingScheduleCollector initializes a new ScalingScheduleCollector.
-func NewScalingScheduleCollector(store Store, defaultScalingWindow time.Duration, now Now, hpa *autoscalingv2.HorizontalPodAutoscaler, config *MetricConfig, interval time.Duration) (*ScalingScheduleCollector, error) {
+func NewScalingScheduleCollector(store Store, defaultScalingWindow time.Duration, rampSteps int, now Now, hpa *autoscalingv2.HorizontalPodAutoscaler, config *MetricConfig, interval time.Duration) (*ScalingScheduleCollector, error) {
 	return &ScalingScheduleCollector{
 		scalingScheduleCollector{
 			store:                store,
@@ -162,12 +167,13 @@ func NewScalingScheduleCollector(store Store, defaultScalingWindow time.Duration
 			interval:             interval,
 			config:               *config,
 			defaultScalingWindow: defaultScalingWindow,
+			rampSteps:            rampSteps,
 		},
 	}, nil
 }

 // NewClusterScalingScheduleCollector initializes a new ScalingScheduleCollector.
-func NewClusterScalingScheduleCollector(store Store, defaultScalingWindow time.Duration, now Now, hpa *autoscalingv2.HorizontalPodAutoscaler, config *MetricConfig, interval time.Duration) (*ClusterScalingScheduleCollector, error) {
+func NewClusterScalingScheduleCollector(store Store, defaultScalingWindow time.Duration, rampSteps int, now Now, hpa *autoscalingv2.HorizontalPodAutoscaler, config *MetricConfig, interval time.Duration) (*ClusterScalingScheduleCollector, error) {
 	return &ClusterScalingScheduleCollector{
 		scalingScheduleCollector{
 			store:                store,
@@ -178,6 +184,7 @@ func NewClusterScalingScheduleCollector(store Store, defaultScalingWindow time.D
 			interval:             interval,
 			config:               *config,
 			defaultScalingWindow: defaultScalingWindow,
+			rampSteps:            rampSteps,
 		},
 	}, nil
 }
@@ -196,7 +203,7 @@ func (c *ScalingScheduleCollector) GetMetrics() ([]CollectedMetric, error) {
 	if !ok {
 		return nil, ErrNotScalingScheduleFound
 	}
-	return calculateMetrics(scalingSchedule.Spec, c.defaultScalingWindow, c.now(), c.objectReference, c.metric)
+	return calculateMetrics(scalingSchedule.Spec, c.defaultScalingWindow, c.rampSteps, c.now(), c.objectReference, c.metric)
 }

 // GetMetrics is the main implementation for collector.Collector interface
@@ -229,7 +236,7 @@ func (c *ClusterScalingScheduleCollector) GetMetrics() ([]CollectedMetric, error
 		clusterScalingSchedule = v1.ClusterScalingSchedule(*scalingSchedule)
 	}

-	return calculateMetrics(clusterScalingSchedule.Spec, c.defaultScalingWindow, c.now(), c.objectReference, c.metric)
+	return calculateMetrics(clusterScalingSchedule.Spec, c.defaultScalingWindow, c.rampSteps, c.now(), c.objectReference, c.metric)
 }

 // Interval returns the interval at which the collector should run.
@@ -242,7 +249,7 @@ func (c *ClusterScalingScheduleCollector) Interval() time.Duration {
 	return c.interval
 }

-func calculateMetrics(spec v1.ScalingScheduleSpec, defaultScalingWindow time.Duration, now time.Time, objectReference custom_metrics.ObjectReference, metric autoscalingv2.MetricIdentifier) ([]CollectedMetric, error) {
+func calculateMetrics(spec v1.ScalingScheduleSpec, defaultScalingWindow time.Duration, rampSteps int, now time.Time, objectReference custom_metrics.ObjectReference, metric autoscalingv2.MetricIdentifier) ([]CollectedMetric, error) {
 	scalingWindowDuration := defaultScalingWindow
 	if spec.ScalingWindowDurationMinutes != nil {
 		scalingWindowDuration = time.Duration(*spec.ScalingWindowDurationMinutes) * time.Minute
@@ -285,7 +292,7 @@ func calculateMetrics(spec v1.ScalingScheduleSpec, defaultScalingWindow time.Dur
 						parsedStartTime.Nanosecond(),
 						location,
 					)
-					value = maxInt64(value, valueForEntry(now, scheduledTime, schedule.Duration(), scalingWindowDuration, schedule.Value))
+					value = maxInt64(value, valueForEntry(now, scheduledTime, schedule.Duration(), scalingWindowDuration, rampSteps, schedule.Value))
 					break
 				}
 			}
@@ -295,7 +302,7 @@ func calculateMetrics(spec v1.ScalingScheduleSpec, defaultScalingWindow time.Dur
 				return nil, ErrInvalidScheduleDate
 			}

-			value = maxInt64(value, valueForEntry(now, scheduledTime, schedule.Duration(), scalingWindowDuration, schedule.Value))
+			value = maxInt64(value, valueForEntry(now, scheduledTime, schedule.Duration(), scalingWindowDuration, rampSteps, schedule.Value))
 		}
 	}

@@ -313,7 +320,7 @@ func calculateMetrics(spec v1.ScalingScheduleSpec, defaultScalingWindow time.Dur
 	}, nil
 }

-func valueForEntry(timestamp time.Time, startTime time.Time, entryDuration time.Duration, scalingWindowDuration time.Duration, value int64) int64 {
+func valueForEntry(timestamp time.Time, startTime time.Time, entryDuration time.Duration, scalingWindowDuration time.Duration, rampSteps int, value int64) int64 {
 	scaleUpStart := startTime.Add(-scalingWindowDuration)
 	endTime := startTime.Add(entryDuration)
 	scaleUpEnd := endTime.Add(scalingWindowDuration)
@@ -322,19 +329,28 @@ func valueForEntry(timestamp time.Time, startTime time.Time, entryDuration time.
 		return value
 	}
 	if between(timestamp, scaleUpStart, startTime) {
-		return scaledValue(timestamp, scaleUpStart, scalingWindowDuration, value)
+		return scaledValue(timestamp, scaleUpStart, scalingWindowDuration, rampSteps, value)
 	}
 	if between(timestamp, endTime, scaleUpEnd) {
-		return scaledValue(scaleUpEnd, timestamp, scalingWindowDuration, value)
+		return scaledValue(scaleUpEnd, timestamp, scalingWindowDuration, rampSteps, value)
 	}
 	return 0
 }

-func scaledValue(timestamp time.Time, startTime time.Time, scalingWindowDuration time.Duration, value int64) int64 {
+// The HPA has a rule to do not scale up or down if the change in the
+// metric is less than 10% (by default) of the current value. We will
+// use buckets of time using the floor of each as the returned metric.
+// Any config greater or equal to 10 buckets must guarantee changes
+// bigger than 10%.
+func scaledValue(timestamp time.Time, startTime time.Time, scalingWindowDuration time.Duration, rampSteps int, value int64) int64 {
 	if scalingWindowDuration == 0 {
 		return 0
 	}
-	return int64(math.Ceil(math.Abs(float64(timestamp.Sub(startTime))) / float64(scalingWindowDuration) * float64(value)))
+
+	steps := float64(rampSteps)
+
+	requiredPercentage := math.Abs(float64(timestamp.Sub(startTime))) / float64(scalingWindowDuration)
+	return int64(math.Floor(requiredPercentage*steps) * (float64(value) / steps))
 }

 func between(timestamp, start, end time.Time) bool {
@@ -15,6 +15,7 @@ import (
 const (
 	hHMMFormat                   = "15:04"
 	defaultScalingWindowDuration = 1 * time.Minute
+	defaultRampSteps             = 10
 )

 type schedule struct {
@@ -48,6 +49,7 @@ func TestScalingScheduleCollector(t *testing.T) {
 		scalingWindowDurationMinutes *int64
 		expectedValue                int64
 		err                          error
+		rampSteps                    int
 	}{
 		{
 			msg: "Return the right value for one time config",
@@ -86,7 +88,7 @@ func TestScalingScheduleCollector(t *testing.T) {
 			expectedValue: 100,
 		},
 		{
-			msg: "Return the scaled value (67) for one time config - 20 seconds before starting",
+			msg: "Return the scaled value (60) for one time config - 20 seconds before starting",
 			schedules: []schedule{
 				{
 					date:     nowTime.Add(time.Second * 20).Format(time.RFC3339),
@@ -95,10 +97,10 @@ func TestScalingScheduleCollector(t *testing.T) {
 					value:    100,
 				},
 			},
-			expectedValue: 67,
+			expectedValue: 60,
 		},
 		{
-			msg: "Return the scaled value (67) for one time config - 20 seconds after",
+			msg: "Return the scaled value (60) for one time config - 20 seconds after",
 			schedules: []schedule{
 				{
 					date:     nowTime.Add(-time.Minute * 45).Add(-time.Second * 20).Format(time.RFC3339),
@@ -107,10 +109,35 @@ func TestScalingScheduleCollector(t *testing.T) {
 					value:    100,
 				},
 			},
-			expectedValue: 67,
+			expectedValue: 60,
 		},
 		{
-			msg:                          "Return the scaled value (95) for one time config with a custom scaling window - 30 seconds before starting",
+			msg: "10 steps (default) return 90% of the metric, even 1 second before",
+			schedules: []schedule{
+				{
+					date:     nowTime.Add(time.Second * 1).Format(time.RFC3339),
+					kind:     "OneTime",
+					duration: 45,
+					value:    100,
+				},
+			},
+			expectedValue: 90,
+		},
+		{
+			msg: "5 steps return 80% of the metric, even 1 second before",
+			schedules: []schedule{
+				{
+					date:     nowTime.Add(time.Second * 1).Format(time.RFC3339),
+					kind:     "OneTime",
+					duration: 45,
+					value:    100,
+				},
+			},
+			expectedValue: 80,
+			rampSteps:     5,
+		},
+		{
+			msg:                          "Return the scaled value (90) for one time config with a custom scaling window - 30 seconds before starting",
 			scalingWindowDurationMinutes: &tenMinutes,
 			schedules: []schedule{
 				{
@@ -120,10 +147,10 @@ func TestScalingScheduleCollector(t *testing.T) {
 					value:    100,
 				},
 			},
-			expectedValue: 95,
+			expectedValue: 90,
 		},
 		{
-			msg:                          "Return the scaled value (95) for one time config with a custom scaling window - 30 seconds after",
+			msg:                          "Return the scaled value (90) for one time config with a custom scaling window - 30 seconds after",
 			scalingWindowDurationMinutes: &tenMinutes,
 			schedules: []schedule{
 				{
@@ -133,7 +160,7 @@ func TestScalingScheduleCollector(t *testing.T) {
 					value:    100,
 				},
 			},
-			expectedValue: 95,
+			expectedValue: 90,
 		},
 		{
 			msg: "Return the default value (0) for one time config not started yet (20 minutes before)",
@@ -471,17 +498,22 @@ func TestScalingScheduleCollector(t *testing.T) {
 			scalingScheduleName := "my_scaling_schedule"
 			namespace := "default"

+			rampSteps := tc.rampSteps
+			if rampSteps == 0 {
+				rampSteps = defaultRampSteps
+			}
+
 			schedules := getSchedules(tc.schedules)
 			store := newMockStore(scalingScheduleName, namespace, tc.scalingWindowDurationMinutes, schedules)
-			plugin, err := NewScalingScheduleCollectorPlugin(store, now, defaultScalingWindowDuration)
+			plugin, err := NewScalingScheduleCollectorPlugin(store, now, defaultScalingWindowDuration, rampSteps)
 			require.NoError(t, err)

 			clusterStore := newClusterMockStore(scalingScheduleName, tc.scalingWindowDurationMinutes, schedules)
-			clusterPlugin, err := NewClusterScalingScheduleCollectorPlugin(clusterStore, now, defaultScalingWindowDuration)
+			clusterPlugin, err := NewClusterScalingScheduleCollectorPlugin(clusterStore, now, defaultScalingWindowDuration, rampSteps)
 			require.NoError(t, err)

 			clusterStoreFirstRun := newClusterMockStoreFirstRun(scalingScheduleName, tc.scalingWindowDurationMinutes, schedules)
-			clusterPluginFirstRun, err := NewClusterScalingScheduleCollectorPlugin(clusterStoreFirstRun, now, defaultScalingWindowDuration)
+			clusterPluginFirstRun, err := NewClusterScalingScheduleCollectorPlugin(clusterStoreFirstRun, now, defaultScalingWindowDuration, rampSteps)
 			require.NoError(t, err)

 			hpa := makeScalingScheduleHPA(namespace, scalingScheduleName)
@@ -549,14 +581,14 @@ func TestScalingScheduleObjectNotPresentReturnsError(t *testing.T) {
 		make(map[string]interface{}),
 		getByKeyFn,
 	}
-	plugin, err := NewScalingScheduleCollectorPlugin(store, time.Now, defaultScalingWindowDuration)
+	plugin, err := NewScalingScheduleCollectorPlugin(store, time.Now, defaultScalingWindowDuration, defaultRampSteps)
 	require.NoError(t, err)

 	clusterStore := mockStore{
 		make(map[string]interface{}),
 		getByKeyFn,
 	}
-	clusterPlugin, err := NewClusterScalingScheduleCollectorPlugin(clusterStore, time.Now, defaultScalingWindowDuration)
+	clusterPlugin, err := NewClusterScalingScheduleCollectorPlugin(clusterStore, time.Now, defaultScalingWindowDuration, defaultRampSteps)
 	require.NoError(t, err)

 	hpa := makeScalingScheduleHPA("namespace", "scalingScheduleName")
@@ -611,10 +643,10 @@ func TestReturnsErrorWhenStoreDoes(t *testing.T) {
 		},
 	}

-	plugin, err := NewScalingScheduleCollectorPlugin(store, time.Now, defaultScalingWindowDuration)
+	plugin, err := NewScalingScheduleCollectorPlugin(store, time.Now, defaultScalingWindowDuration, defaultRampSteps)
 	require.NoError(t, err)

-	clusterPlugin, err := NewClusterScalingScheduleCollectorPlugin(store, time.Now, defaultScalingWindowDuration)
+	clusterPlugin, err := NewClusterScalingScheduleCollectorPlugin(store, time.Now, defaultScalingWindowDuration, defaultRampSteps)
 	require.NoError(t, err)

 	hpa := makeScalingScheduleHPA("namespace", "scalingScheduleName")
@@ -269,7 +269,7 @@ func (p *HPAProvider) collectMetrics(ctx context.Context) {

 // GetMetricByName gets a single metric by name.
 func (p *HPAProvider) GetMetricByName(name types.NamespacedName, info provider.CustomMetricInfo, metricSelector labels.Selector) (*custom_metrics.MetricValue, error) {
-	metric := p.metricStore.GetMetricsByName(name, info)
+	metric := p.metricStore.GetMetricsByName(name, info, metricSelector)
 	if metric == nil {
 		return nil, provider.NewMetricNotFoundForError(info.GroupResource, info.Metric, name.Name)
 	}
@@ -279,7 +279,7 @@ func (p *HPAProvider) GetMetricByName(name types.NamespacedName, info provider.C
 // GetMetricBySelector returns metrics for namespaced resources by
 // label selector.
 func (p *HPAProvider) GetMetricBySelector(namespace string, selector labels.Selector, info provider.CustomMetricInfo, metricSelector labels.Selector) (*custom_metrics.MetricValueList, error) {
-	return p.metricStore.GetMetricsBySelector(namespace, selector, info), nil
+	return p.metricStore.GetMetricsBySelector(objectNamespace(namespace), selector, info), nil
 }

 // ListAllMetrics list all available metrics from the provicer.
@@ -288,7 +288,7 @@ func (p *HPAProvider) ListAllMetrics() []provider.CustomMetricInfo {
 }

 func (p *HPAProvider) GetExternalMetric(namespace string, metricSelector labels.Selector, info provider.ExternalMetricInfo) (*external_metrics.ExternalMetricValueList, error) {
-	return p.metricStore.GetExternalMetric(namespace, metricSelector, info)
+	return p.metricStore.GetExternalMetric(objectNamespace(namespace), metricSelector, info)
 }

 func (p *HPAProvider) ListAllExternalMetrics() []provider.ExternalMetricInfo {
@@ -128,7 +128,8 @@ func NewCommandStartAdapterServer(stopCh <-chan struct{}) *cobra.Command {
 	flags.DurationVar(&o.GCInterval, "garbage-collector-interval", 10*time.Minute, "Interval to clean up metrics that are stored in in-memory cache.")
 	flags.BoolVar(&o.ScalingScheduleMetrics, "scaling-schedule", o.ScalingScheduleMetrics, ""+
 		"whether to enable time-based ScalingSchedule metrics")
-	flags.DurationVar(&o.DefaultScheduledScalingWindow, "scaling-schedule-default-scaling-window", 10*time.Minute, "Default scale-up/scale-down window duration for scheduled metrics")
+	flags.DurationVar(&o.DefaultScheduledScalingWindow, "scaling-schedule-default-scaling-window", 10*time.Minute, "Default rampup and rampdown window duration for ScalingSchedules")
+	flags.IntVar(&o.RampSteps, "scaling-schedule-ramp-steps", 10, "Number of steps used to rampup and rampdown ScalingSchedules. It's used to guarantee won't avoid reaching the max scaling due to the 10% minimum change rule.")
 	return cmd
 }

@@ -294,7 +295,7 @@ func (o AdapterServerOptions) RunCustomMetricsAdapterServer(stopCh <-chan struct
 		)
 		go reflector.Run(ctx.Done())

-		clusterPlugin, err := collector.NewClusterScalingScheduleCollectorPlugin(clusterScalingSchedulesStore, time.Now, o.DefaultScheduledScalingWindow)
+		clusterPlugin, err := collector.NewClusterScalingScheduleCollectorPlugin(clusterScalingSchedulesStore, time.Now, o.DefaultScheduledScalingWindow, o.RampSteps)
 		if err != nil {
 			return fmt.Errorf("unable to create ClusterScalingScheduleCollector plugin: %v", err)
 		}
@@ -303,7 +304,7 @@ func (o AdapterServerOptions) RunCustomMetricsAdapterServer(stopCh <-chan struct
 			return fmt.Errorf("failed to register ClusterScalingSchedule object collector plugin: %v", err)
 		}

-		plugin, err := collector.NewScalingScheduleCollectorPlugin(scalingSchedulesStore, time.Now, o.DefaultScheduledScalingWindow)
+		plugin, err := collector.NewScalingScheduleCollectorPlugin(scalingSchedulesStore, time.Now, o.DefaultScheduledScalingWindow, o.RampSteps)
 		if err != nil {
 			return fmt.Errorf("unable to create ScalingScheduleCollector plugin: %v", err)
 		}
@@ -429,6 +430,8 @@ type AdapterServerOptions struct {
 	GCInterval time.Duration
 	// Time-based scaling based on the CRDs ScheduleScaling and ClusterScheduleScaling.
 	ScalingScheduleMetrics bool
-	// Default scale-up/scale-down window duration for scheduled metrics
+	// Default ramp-up/ramp-down window duration for scheduled metrics
 	DefaultScheduledScalingWindow time.Duration
+	// Number of steps utilized during the rampup and rampdown for scheduled metrics
+	RampSteps int
 }
Author	SHA1	Message	Date
Mikkel Oscar Lyderik Larsen	10dc42b9e9	Merge pull request #401 from zalando-incubator/update-deps-02-14 Update dependencies	2022-02-14 15:05:19 +01:00
Mikkel Oscar Lyderik Larsen	20db592a62	Use cdp-runtime for build (#398 ) Signed-off-by: Mikkel Oscar Lyderik Larsen <mikkel.larsen@zalando.de>	2022-02-14 14:41:57 +01:00
Mikkel Oscar Lyderik Larsen	d0a3ea1934	Update dependencies Signed-off-by: Mikkel Oscar Lyderik Larsen <mikkel.larsen@zalando.de>	2022-02-14 14:36:06 +01:00
Mikkel Oscar Lyderik Larsen	837e7b9c5d	Update deps 2022 02 (#391 ) * Update example Signed-off-by: Mikkel Oscar Lyderik Larsen <mikkel.larsen@zalando.de> * Update dependencies 2022-02 Signed-off-by: Mikkel Oscar Lyderik Larsen <mikkel.larsen@zalando.de>	2022-02-01 18:13:48 +01:00
Katyanna Moura	9d8359b580	Merge pull request #380 from zalando-incubator/fix-metrics-map Fix panic when trying to add key to nil map	2022-01-12 11:36:53 +01:00
Katyanna Moura	71a8e99d1f	Fix panic when trying to add key to nil map Bug was introduced trying to solved the following issue: > When scenarios where two HPAs reference the same object but have different metric calculation (e.g. ingresses with different weights), kube-metrics-adapter calculates the two metrics but always overrides one of the metrics when saving it to the store. This commit fixes a issue where without the added `return` the system was continuing in an invalid states where `labels2metric` wasn't properly initialized, causing the system to panic. Signed-off-by: Katyanna Moura <amelie.kn@gmail.com>	2022-01-12 11:29:38 +01:00
Katyanna Moura	4d4c70c553	Fix test filename to match related file Having the test filename prefix exactly as the tested file name helps systems to identify the correlation. Signed-off-by: Katyanna Moura <amelie.kn@gmail.com>	2022-01-12 11:29:38 +01:00
Katyanna Moura	2ccd6903d9	Merge pull request #376 from zalando-incubator/custommetricsselector Use labels hash in the custom metrics store and metrics store 💅	2021-11-25 10:24:44 +01:00
Jonathan Juares Beber	f58db31f98	Define stronger types for metrics store The metrics store, both the custom and external one, make heavy usage of maps of strings to strings. This leads to a presumable confusing or at least hard to read codebase. This commit tries to increase readability and maintainability of the metric stores by using custom types. This commit has no effect in performance or functionality, it's pure aesthetics. Co-authored-by: Jonathan Juares Beber <jonathanbeber@gmail.com> Signed-off-by: Katyanna Moura <amelie.kn@gmail.com>	2021-11-24 17:41:03 +01:00
Jonathan Juares Beber	0bf8f5dd0f	Use labels hash in the custom metrics store The current implementation of the metrics store for custom metrics uses just the object name as key. When scenarios where two HPAs reference the same object but have different metric calculation (e.g. ingresses with different weights), kube-metrics-adapter calculates the two metrics but always overrides one of the metrics when saving it to the store. This commit implements the use of a labels hash in the custom metrics store. This way, metrics are identified not just by the referenced object, but also by its selectors. Co-authored-by: Katyanna Moura <amelie.kn@gmail.com> Signed-off-by: Jonathan Juares Beber <jonathanbeber@gmail.com>	2021-11-23 18:11:18 +01:00
Jonathan Juares Beber	e600557636	Merge pull request #375 from zalando-incubator/kube1.21 Upgrade Kubernetes and its friends to 1.21	2021-10-28 10:35:11 +02:00
Jonathan Juares Beber	0f06db7cdf	Upgrade Kubernetes and its friends to 1.21 This commit upgrades all the Kubernetes dependencies to 1.21.5. It includes regenerated clients and specs. Signed-off-by: Jonathan Juares Beber <jonathanbeber@gmail.com>	2021-10-27 20:58:18 +02:00
Jonathan Juares Beber	1c9038b2cc	Merge pull request #374 from zalando-incubator/configurable-buckets Make the number of ramp steps configurable	2021-10-25 10:21:32 +02:00
Jonathan Juares Beber	fd4ead837e	Make the number of ramp steps configurable In #371 we introduced steps to make the scaling up possible even when the HPA forces a 10% change. The problem is that 10% might not be sufficient for some specific scaling scenarios. For example, a an application targeting 12 pods and using a ScalingSchedule with the value of 10000 to achieve that, will require a target of 833. With 10 ramp steps the 90% bucket will return a metric of 9000 and the HPA calculates (9000/833) 10.8 pods, rounding to 11 pods. Once the metric reaches the time to return 100% it will won't be effective, since the change of the current number of pods (11) and the desired one (12) is less than 10%. This commit does not try to tackle this problem completely, since the 10% rule is not fixed, might change among different clusters and is also dependent on the value given to each ScalingSchedule. Therefore, this commit makes the number of ramp steps configurable via the `--scaling-schedule-ramp-steps` config flag, defaulting to 10. Signed-off-by: Jonathan Juares Beber <jonathanbeber@gmail.com>	2021-10-22 15:35:11 +02:00
Mikkel Oscar Lyderik Larsen	f46f801811	Merge pull request #373 from zalando-incubator/json-path-array Handle more complex array in json path	2021-10-19 10:04:04 +02:00
Mikkel Oscar Lyderik Larsen	4acdf72ef7	Handle more complex array in json path Signed-off-by: Mikkel Oscar Lyderik Larsen <mikkel.larsen@zalando.de>	2021-10-14 09:34:15 +02:00
Jonathan Juares Beber	e04cd10bfc	Merge pull request #371 from zalando-incubator/scaling-chunks Use 10 buckets on ScalingSchedule ramp-up/down	2021-10-01 10:39:57 +02:00
Jonathan Juares Beber	8fe330941a	Use 10 buckets on ScalingSchedule ramp-up/down The HPA has a feature to do not scale up and down when the change in the metric is less than 10%: > We'll skip scaling if the ratio is sufficiently close to 1.0 (within a > globally-configurable tolerance, from the > `--horizontal-pod-autoscaler-tolerance` flag, which defaults to 0.1. It could lead to pods scaling up to 10% less than the target for ScalingSchedules and then not scaling to the actual value if the metric calculated before was less than 10% of the target. This commit uses 10 fixed buckets for scaling, this way we know the metric returned during a scaling event is at least 10% more than a previous one calculated during the period of ramp up. The same is valid for the scaling down during a ramp-down Signed-off-by: Jonathan Juares Beber <jonathanbeber@gmail.com>	2021-09-30 19:01:59 +02:00