Merge pull request #636 from zalando-incubator/nakadi-collector

Add support for scaling based on Nakadi

README.md

@@ -702,6 +702,95 @@ you need to define a `key` or other `tag` with a "star" query syntax like
 metric label definitions. If both annotations and corresponding label is
 defined, then the annotation takes precedence.
 
+
+## Nakadi collector
+
+The Nakadi collector allows scaling based on [Nakadi](https://nakadi.io/)
+ Subscription API stats metrics `consumer_lag_seconds` or `unconsumed_events`.
+
+### Supported metrics
+
+| Metric Type            | Description                                                                 | Type     | K8s Versions |
+|------------------------|-----------------------------------------------------------------------------|----------|--------------|
+| `unconsumed-events`    | Scale based on number of unconsumed events for a Nakadi subscription        | External | `>=1.24`     |
+| `consumer-lag-seconds` | Scale based on number of max consumer lag seconds for a Nakadi subscription | External | `>=1.24`     |
+| ------------ | ------- | -- | -- |
+| `unconsumed-events` | Scale based on number of unconsumed events for a Nakadi Subscription | External | `>=1.24` |
+| `consumer-lag-seconds` | Scale based on number of max consumer lag seconds for a Nakadi Subscription | External | `>=1.24` |
+
+```yaml
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  name: myapp-hpa
+  annotations:
+    # metric-config.<metricType>.<metricName>.<collectorType>/<configKey>
+    metric-config.external.my-nakadi-consumer.nakadi/interval: "60s" # optional
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: custom-metrics-consumer
+  minReplicas: 0
+  maxReplicas: 8 # should match number of partitions for the event type
+  metrics:
+  - type: External
+    external:
+      metric:
+          name: my-nakadi-consumer
+          selector:
+            matchLabels:
+              type: nakadi
+              subscription-id: "708095f6-cece-4d02-840e-ee488d710b29"
+              metric-type: "consumer-lag-seconds|unconsumed-events"
+      target:
+        # value is compatible with the consumer-lag-seconds metric type.
+        # It describes the amount of consumer lag in seconds before scaling
+        # additionally up.
+        # if an event-type has multiple partitions the value of
+        # consumer-lag-seconds is the max of all the partitions.
+        value: "600" # 10m
+        # averageValue is compatible with unconsumed-events metric type.
+        # This means for every 30 unconsumed events a pod is added.
+        # unconsumed-events is the sum of of unconsumed_events over all
+        # partitions.
+        averageValue: "30"
+        type: AverageValue
+```
+
+The `subscription-id` is the Subscription ID of the relevant consumer. The
+`metric-type` indicates whether to scale on `consumer-lag-seconds` or
+`unconsumed-events` as outlined below.
+
+`unconsumed-events` - is the total number of unconsumed events over all
+partitions. When using this `metric-type` you should also use the target
+`averageValue` which indicates the number of events which can be handled per
+pod. To best estimate the number of events per pods, you need to understand the
+average time for processing an event as well as the rate of events.
+
+*Example*: You have an event type producing 100 events per second between 00:00
+and 08:00. Between 08:01 to 23:59 it produces 400 events per second.
+Let's assume that on average a single pod can consume 100 events per second,
+then we can define 100 as `averageValue` and the HPA would scale to 1 between
+00:00 and 08:00, and scale to 4 between 08:01 and 23:59. If there for some
+reason is a short spike of 800 events per second, then it would scale to 8 pods
+to process those events until the rate goes down again.
+
+`consumer-lag-seconds` -  describes the age of the oldest unconsumed event for
+a subscription. If the event type has multiple partitions the lag is defined as
+the max age over all partitions. When using this `metric-type` you should use
+the target `value` to indicate the max lag (in seconds) before the HPA should
+scale.
+
+*Example*: You have a subscription with a defined SLO of "99.99 of events are
+consumed within 30 min.". In this case you can define a target `value` of e.g.
+20 min. (1200s) (to include a safety buffer) such that the HPA only scales up
+from 1 to 2 if the target of 20 min. is breached and it needs to work faster
+with more consumers.
+For this case you should also account for the average time for processing an
+event when defining the target.
+
+
 ## HTTP Collector
 
 The http collector allows collecting metrics from an external endpoint specified in the HPA.

pkg/collector/nakadi_collector.go

@@ -0,0 +1,120 @@
+package collector
+
+import (
+	"context"
+	"fmt"
+	"time"
+
+	"github.com/zalando-incubator/kube-metrics-adapter/pkg/nakadi"
+	autoscalingv2 "k8s.io/api/autoscaling/v2"
+	"k8s.io/apimachinery/pkg/api/resource"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/metrics/pkg/apis/external_metrics"
+)
+
+const (
+	// NakadiMetricType defines the metric type for metrics based on Nakadi
+	// subscriptions.
+	NakadiMetricType                   = "nakadi"
+	nakadiSubscriptionIDKey            = "subscription-id"
+	nakadiMetricTypeKey                = "metric-type"
+	nakadiMetricTypeConsumerLagSeconds = "consumer-lag-seconds"
+	nakadiMetricTypeUnconsumedEvents   = "unconsumed-events"
+)
+
+// NakadiCollectorPlugin defines a plugin for creating collectors that can get
+// unconsumed events from Nakadi.
+type NakadiCollectorPlugin struct {
+	nakadi nakadi.Nakadi
+}
+
+// NewNakadiCollectorPlugin initializes a new NakadiCollectorPlugin.
+func NewNakadiCollectorPlugin(nakadi nakadi.Nakadi) (*NakadiCollectorPlugin, error) {
+	return &NakadiCollectorPlugin{
+		nakadi: nakadi,
+	}, nil
+}
+
+// NewCollector initializes a new Nakadi collector from the specified HPA.
+func (c *NakadiCollectorPlugin) NewCollector(hpa *autoscalingv2.HorizontalPodAutoscaler, config *MetricConfig, interval time.Duration) (Collector, error) {
+	return NewNakadiCollector(c.nakadi, hpa, config, interval)
+}
+
+// NakadiCollector defines a collector that is able to collect metrics from
+// Nakadi.
+type NakadiCollector struct {
+	nakadi           nakadi.Nakadi
+	interval         time.Duration
+	subscriptionID   string
+	nakadiMetricType string
+	metric           autoscalingv2.MetricIdentifier
+	metricType       autoscalingv2.MetricSourceType
+	namespace        string
+}
+
+// NewNakadiCollector initializes a new NakadiCollector.
+func NewNakadiCollector(nakadi nakadi.Nakadi, hpa *autoscalingv2.HorizontalPodAutoscaler, config *MetricConfig, interval time.Duration) (*NakadiCollector, error) {
+	if config.Metric.Selector == nil {
+		return nil, fmt.Errorf("selector for nakadi is not specified")
+	}
+
+	subscriptionID, ok := config.Config[nakadiSubscriptionIDKey]
+	if !ok {
+		return nil, fmt.Errorf("subscription-id not specified on metric")
+	}
+
+	metricType, ok := config.Config[nakadiMetricTypeKey]
+	if !ok {
+		return nil, fmt.Errorf("metric-type not specified on metric")
+	}
+
+	if metricType != nakadiMetricTypeConsumerLagSeconds && metricType != nakadiMetricTypeUnconsumedEvents {
+		return nil, fmt.Errorf("metric-type must be either '%s' or '%s', was '%s'", nakadiMetricTypeConsumerLagSeconds, nakadiMetricTypeUnconsumedEvents, metricType)
+	}
+
+	return &NakadiCollector{
+		nakadi:           nakadi,
+		interval:         interval,
+		subscriptionID:   subscriptionID,
+		nakadiMetricType: metricType,
+		metric:           config.Metric,
+		metricType:       config.Type,
+		namespace:        hpa.Namespace,
+	}, nil
+}
+
+// GetMetrics returns a list of collected metrics for the Nakadi subscription ID.
+func (c *NakadiCollector) GetMetrics() ([]CollectedMetric, error) {
+	var value int64
+	var err error
+	switch c.nakadiMetricType {
+	case nakadiMetricTypeConsumerLagSeconds:
+		value, err = c.nakadi.ConsumerLagSeconds(context.TODO(), c.subscriptionID)
+		if err != nil {
+			return nil, err
+		}
+	case nakadiMetricTypeUnconsumedEvents:
+		value, err = c.nakadi.UnconsumedEvents(context.TODO(), c.subscriptionID)
+		if err != nil {
+			return nil, err
+		}
+	}
+
+	metricValue := CollectedMetric{
+		Namespace: c.namespace,
+		Type:      c.metricType,
+		External: external_metrics.ExternalMetricValue{
+			MetricName:   c.metric.Name,
+			MetricLabels: c.metric.Selector.MatchLabels,
+			Timestamp:    metav1.Now(),
+			Value:        *resource.NewQuantity(value, resource.DecimalSI),
+		},
+	}
+
+	return []CollectedMetric{metricValue}, nil
+}
+
+// Interval returns the interval at which the collector should run.
+func (c *NakadiCollector) Interval() time.Duration {
+	return c.interval
+}

pkg/nakadi/nakadi.go

@@ -0,0 +1,124 @@
+package nakadi
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+)
+
+// Nakadi defines an interface for talking to the Nakadi API.
+type Nakadi interface {
+	ConsumerLagSeconds(ctx context.Context, subscriptionID string) (int64, error)
+	UnconsumedEvents(ctx context.Context, subscriptionID string) (int64, error)
+}
+
+// Client defines client for interfacing with the Nakadi API.
+type Client struct {
+	nakadiEndpoint string
+	http           *http.Client
+}
+
+// NewNakadiClient initializes a new Nakadi Client.
+func NewNakadiClient(nakadiEndpoint string, client *http.Client) *Client {
+	return &Client{
+		nakadiEndpoint: nakadiEndpoint,
+		http:           client,
+	}
+}
+
+func (c *Client) ConsumerLagSeconds(ctx context.Context, subscriptionID string) (int64, error) {
+	stats, err := c.stats(ctx, subscriptionID)
+	if err != nil {
+		return 0, err
+	}
+
+	var maxConsumerLagSeconds int64
+	for _, eventType := range stats {
+		for _, partition := range eventType.Partitions {
+			maxConsumerLagSeconds = max(maxConsumerLagSeconds, partition.ConsumerLagSeconds)
+		}
+	}
+
+	return maxConsumerLagSeconds, nil
+}
+
+func (c *Client) UnconsumedEvents(ctx context.Context, subscriptionID string) (int64, error) {
+	stats, err := c.stats(ctx, subscriptionID)
+	if err != nil {
+		return 0, err
+	}
+
+	var unconsumedEvents int64
+	for _, eventType := range stats {
+		for _, partition := range eventType.Partitions {
+			unconsumedEvents += partition.UnconsumedEvents
+		}
+	}
+
+	return unconsumedEvents, nil
+}
+
+type statsResp struct {
+	Items []statsEventType `json:"items"`
+}
+
+type statsEventType struct {
+	EventType  string           `json:"event_type"`
+	Partitions []statsPartition `json:"partitions"`
+}
+
+type statsPartition struct {
+	Partiton           string `json:"partition"`
+	State              string `json:"state"`
+	UnconsumedEvents   int64  `json:"unconsumed_events"`
+	ConsumerLagSeconds int64  `json:"consumer_lag_seconds"`
+	StreamID           string `json:"stream_id"`
+	AssignmentType     string `json:"assignment_type"`
+}
+
+// stats returns the Nakadi stats for a given subscription ID.
+//
+// https://nakadi.io/manual.html#/subscriptions/subscription_id/stats_get
+func (c *Client) stats(ctx context.Context, subscriptionID string) ([]statsEventType, error) {
+	endpoint, err := url.Parse(c.nakadiEndpoint)
+	if err != nil {
+		return nil, err
+	}
+
+	endpoint.Path = fmt.Sprintf("/subscriptions/%s/stats", subscriptionID)
+
+	q := endpoint.Query()
+	q.Set("show_time_lag", "true")
+	endpoint.RawQuery = q.Encode()
+
+	resp, err := c.http.Get(endpoint.String())
+	if err != nil {
+		return nil, err
+	}
+	defer resp.Body.Close()
+
+	d, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, err
+	}
+
+	if resp.StatusCode != http.StatusOK {
+		return nil, fmt.Errorf("[nakadi stats] unexpected response code: %d (%s)", resp.StatusCode, string(d))
+	}
+
+	var result statsResp
+	err = json.Unmarshal(d, &result)
+	if err != nil {
+		return nil, err
+	}
+
+	if len(result.Items) == 0 {
+		return nil, errors.New("expected at least 1 event-type, 0 returned")
+	}
+
+	return result.Items, nil
+}