[hpa]only set scaleDown, then with many error message

[jinyingsunny]

Disabled scale, yaml as follow，when pressure test stop it keep pod alive，however there are some error in log

errors in log：

github.com/vesoft-inc/nebula-operator/pkg/controller/autoscaler.(*HorizontalController).stabilizeRecommendationWithBehaviors(0xc000517290, {{0xc000fac6f0, 0x18}, 0x0, 0xc0007339e0, 0x2, 0xa, 0x2, 0x1})
	github.com/vesoft-inc/nebula-operator/pkg/controller/autoscaler/autoscaler.go:713 +0xc5
github.com/vesoft-inc/nebula-operator/pkg/controller/autoscaler.(*HorizontalController).normalizeDesiredReplicasWithBehaviors(0xc000517290, 0xc000c036c0, {0xc000fac6f0?, 0x18?}, 0x2?, 0x1, 0x2?)
	github.com/vesoft-inc/nebula-operator/pkg/controller/autoscaler/autoscaler.go:603 +0x1f4
github.com/vesoft-inc/nebula-operator/pkg/controller/autoscaler.(*HorizontalController).reconcileAutoscaler(0xc000517290, {0x25044a8, 0xc0007338f0}, 0xc000c03380, {0xc000fac6f0, 0x18})
	github.com/vesoft-inc/nebula-operator/pkg/controller/autoscaler/autoscaler.go:340 +0xada
github.com/vesoft-inc/nebula-operator/pkg/controller/autoscaler.(*HorizontalController).Reconcile(0xc000517290, {0x25044a8, 0xc0007338f0}, {{{0xc000c97a86?, 0x0?}, {0xc0010bf200?, 0x40e027?}}})
	github.com/vesoft-inc/nebula-operator/pkg/controller/autoscaler/autoscaler.go:232 +0x49b
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Reconcile(0x25044a8?, {0x25044a8?, 0xc0007338f0?}, {{{0xc000c97a86?, 0x1dbf260?}, {0xc0010bf200?, 0x24ec878?}}})
	sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:118 +0xc8
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler(0xc000245b80, {0x2504400, 0xc00009ad20}, {0x1f91a20?, 0xc0000a19e0?})
	sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:314 +0x377
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem(0xc000245b80, {0x2504400, 0xc00009ad20})
	sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:265 +0x1d9
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2.2()
	sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:226 +0x85
created by sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2
	sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:222 +0x587
E1006 08:08:38.743381       1 controller.go:324] "Reconciler error" err="panic: runtime error: invalid memory address or nil pointer dereference [recovered]" controller="nebulaautoscaler" controllerGroup="autoscaling.nebula-graph.io" controllerKind="NebulaAutoscaler" NebulaAutoscaler="nebula/nebula-autoscaler" namespace="nebula" name="nebula-autoscaler" reconcileID="4c635937-8fa6-4b43-bb7d-2e1add916132"
I1006 08:08:52.587035       1 autoscaler.go:206] Finished reconciling NebulaAutoscaler [nebula/nebula-autoscaler], spendTime: (9.48868ms)
E1006 08:08:52.587121       1 runtime.go:79] Observed a panic: "invalid memory address or nil pointer dereference" (runtime error: invalid memory address or nil pointer dereference)
goroutine 200 [running]:
k8s.io/apimachinery/pkg/util/runtime.logPanic({0x1eb31e0?, 0x3634650})
	k8s.io/[email protected]/pkg/util/runtime/runtime.go:75 +0x99
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Reconcile.func1()
	sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:107 +0xc5
panic({0x1eb31e0, 0x3634650})
	runtime/panic.go:884 +0x213

however when add scale up witch is default value , error message disappear.
autoscale.yaml as follows:

apiVersion: autoscaling.nebula-graph.io/v1alpha1
kind: NebulaAutoscaler
metadata:
  name: nebula-autoscaler
  namespace: nebula
spec:
  nebulaClusterRef:
    name: nebulazcert
  graphdPolicy:
    minReplicas: 2
    maxReplicas: 10
    metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 50
    behavior:
      scaleUp:
        stabilizationWindowSeconds: 0
        policies:
        - type: Percent
          value: 100
          periodSeconds: 15
        - type: Pods
          value: 4
          periodSeconds: 15
      scaleDown:
        selectPolicy: Disabled
  pollingPeriod: 30s

Your Environments (required)

nebula-operator: s-1.18

How To Reproduce(required)

Steps to reproduce the behavior:

1. set autoscale.yaml
2. provide many query, auto scale many pod
3. show auto scale logs： kubectl -n nebula-operator-system logs -f nebula-operator-controller-manager-deployment-759f97f578-fq6v6 -c autoscaler

Expected behavior
without error message

[jinyingsunny]

when set like follower as demo from https://kubernetes.io/zh-cn/docs/tasks/run-application/horizontal-pod-autoscale/ ,still report error

[MegaByte875]

SelectPolicy is a pointer type and has no default value set, I will give it an initial default value.

[MegaByte875]

#319
#321

[jinyingsunny]

验证如下autoscale.yaml配置,已经可以成功写入和生效：
缩容的稳定窗口是2min，过了2min，如果依旧低于平均cpu使用量，则会进行缩容
operator版本: s-1.20

$ cat autoscale.yaml
apiVersion: autoscaling.nebula-graph.io/v1alpha1
kind: NebulaAutoscaler
metadata:
  name: nebula-autoscaler
  namespace: nebula
spec:
  nebulaClusterRef:
    name: nebulazone
  graphdPolicy:
    minReplicas: 2
    maxReplicas: 5
    metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 10
    behavior:
      scaleUp:
        stabilizationWindowSeconds: 0
        policies:
        - type: Percent
          value: 100
          periodSeconds: 15
        - type: Pods
          value: 4
          periodSeconds: 15
        selectPolicy: Max
      scaleDown:
        stabilizationWindowSeconds: 120
        policies:
        - type: Percent
          value: 100
          periodSeconds: 15
  pollingPeriod: 30s

[jinyingsunny]

验证如下autoscale.yaml配置,apply后不再报错，但是有流量时也不会进行扩容

operator版本: s-1.21

apiVersion: autoscaling.nebula-graph.io/v1alpha1
kind: NebulaAutoscaler
metadata:
  name: nebula-autoscaler
  namespace: nebula
spec:
  nebulaClusterRef:
    name: nebulazone
  graphdPolicy:
    minReplicas: 2
    maxReplicas: 5
    metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 10
    behavior:
        scaleDown:
          selectPolicy: Disabled
  pollingPeriod: 30s

qiaolei：解释 scale up 和 scale down必须同时配置才可以生效 且 selectPolicy一定要设置，否则不生效(也没啥报错，就是average cpu的值一直是0)。