Goldilocks makes delay on pods initialization

[bigsem89]

What happened?

after installationg using helm chart all new pods in cluster start up with quite big delay.

What did you expect to happen?

no delays or have ability to exclude some pods.

How can we reproduce this?

uninstallVPA -- Enabling this flag will remove a vpa installation that was previously managed with this chart. It is considered deprecated and will be removed in a later release.

uninstallVPA: false

vpa:

vpa.enabled -- If true, the vpa will be installed as a sub-chart

enabled: true
updater:
enabled: false

metrics-server:

metrics-server.enabled -- If true, the metrics-server will be installed as a sub-chart

enabled: false
apiService:
create: true

image:

image.repository -- Repository for the goldilocks image

repository: us-docker.pkg.dev/fairwinds-ops/oss/goldilocks

image.tag -- The goldilocks image tag to use

tag: v4.10.0

image.pullPolicy -- imagePullPolicy - Highly recommended to leave this as Always

pullPolicy: Always

imagePullSecrets -- A list of image pull secret names to use

imagePullSecrets: []

nameOverride: ""
fullnameOverride: ""

controller:

controller.enabled -- Whether or not to install the controller deployment

enabled: true

controller.revisionHistoryLimit -- Number of old replicasets to retain, default is 10, 0 will garbage-collect old replicasets

revisionHistoryLimit: 10
rbac:
# controller.rbac.create -- If set to true, rbac resources will be created for the controller
create: true
# controller.rbac.enableArgoproj -- If set to true, the clusterrole will give access to argoproj.io resources
enableArgoproj: true
# controller.rbac.extraRules -- Extra rbac rules for the controller clusterrole
extraRules: []
# controller.rbac.extraClusterRoleBindings -- A list of ClusterRoles for which ClusterRoleBindings will be created for the ServiceAccount, if enabled
extraClusterRoleBindings: []
serviceAccount:
# controller.serviceAccount.create -- If true, a service account will be created for the controller. If set to false, you must set controller.serviceAccount.name
create: true
# controller.serviceAccount.name -- The name of an existing service account to use for the controller. Combined with controller.serviceAccount.create
name:

controller.flags -- A map of additional flags to pass to the controller

flags: {}

controller.logVerbosity -- Controller log verbosity. Can be set from 1-10 with 10 being extremely verbose

logVerbosity: "2"

controller.nodeSelector -- Node selector for the controller pod

nodeSelector: {}

controller.tolerations -- Tolerations for the controller pod

tolerations: []

controller.affinity -- Affinity for the controller pods

affinity: {}

controller.topologySpreadConstraints -- Topology spread constraints for the controller pods

topologySpreadConstraints: []

controller.resources -- The resources block for the controller pods

resources:
limits: {}
requests:
cpu: 25m
memory: 256Mi

controller.podSecurityContext -- Defines the podSecurityContext for the controller pod

podSecurityContext:
seccompProfile:
type: RuntimeDefault

controller.securityContext -- The container securityContext for the controller container

securityContext:
readOnlyRootFilesystem: true
allowPrivilegeEscalation: false
runAsNonRoot: true
runAsUser: 10324
capabilities:
drop:
- ALL

deployment:
# controller.deployment.extraVolumeMounts -- Extra volume mounts for the controller container
extraVolumeMounts: []
# controller.deployment.extraVolumes -- Extra volumes for the controller pod
extraVolumes: []
# controller.deployment.annotations -- Extra annotations for the controller deployment
annotations: {}
# controller.deployment.additionalLabels -- Extra labels for the controller deployment
additionalLabels: {}

# controller.deployment.podAnnotations -- Extra annotations for the controller pod
podAnnotations: {}

dashboard:

dashboard.basePath -- Path on which the dashboard is served. Defaults to /

basePath: null

dashboard.enabled -- If true, the dashboard component will be installed

enabled: true

dashboard.revisionHistoryLimit -- Number of old replicasets to retain, default is 10, 0 will garbage-collect old replicasets

revisionHistoryLimit: 0

dashboard.replicaCount -- Number of dashboard pods to run

replicaCount: 1
service:
# dashboard.service.type -- The type of the dashboard service
type: ClusterIP
# dashboard.service.port -- The port to run the dashboard service on
port: 80
# dashboard.service.annotations -- Extra annotations for the dashboard service
annotations: {}

dashboard.flags -- A map of additional flags to pass to the dashboard

flags: {}

dashboard.logVerbosity -- Dashboard log verbosity. Can be set from 1-10 with 10 being extremely verbose

logVerbosity: "2"

dashboard.excludeContainers -- Container names to exclude from displaying in the Goldilocks dashboard

excludeContainers: "linkerd-proxy,istio-proxy"
rbac:
# dashboard.rbac.create -- If set to true, rbac resources will be created for the dashboard
create: true
# dashboard.rbac.enableArgoproj -- If set to true, the clusterrole will give access to argoproj.io resources
enableArgoproj: true
serviceAccount:
# dashboard.serviceAccount.create -- If true, a service account will be created for the dashboard. If set to false, you must set dashboard.serviceAccount.name
create: true
# dashboard.serviceAccount.name -- The name of an existing service account to use for the controller. Combined with dashboard.serviceAccount.create
name:

deployment:
# dashboard.deployment.annotations -- Extra annotations for the dashboard deployment
annotations: {}
# dashboard.deployment.additionalLabels -- Extra labels for the dashboard deployment
additionalLabels: {}
# dashboard.deployment.extraVolumeMounts -- Extra volume mounts for the dashboard container
extraVolumeMounts: []
# dashboard.deployment.extraVolumes -- Extra volumes for the dashboard pod
extraVolumes: []

# dashboard.deployment.podAnnotations -- Extra annotations for the dashboard pod
podAnnotations: {}

ingress:
# dashboard.ingress.enabled -- Enables an ingress object for the dashboard.
enabled: false

# dashboard.ingress.ingressClassName -- From Kubernetes 1.18+ this field is supported in case your ingress controller supports it. When set, you do not need to add the ingress class as annotation.
ingressClassName:
annotations: {}
  # kubernetes.io/ingress.class: nginx
  # kubernetes.io/tls-acme: "true"
hosts:
  - host: chart-example.local
    paths:
      - path: /
        type: ImplementationSpecific

tls: []
#  - secretName: chart-example-tls
#    hosts:
#      - chart-example.local

dashboard.resources -- A resources block for the dashboard.

resources:
limits: {}
requests:
cpu: 25m
memory: 256Mi

dashboard.podSecurityContext -- Defines the podSecurityContext for the dashboard pod

podSecurityContext:
seccompProfile:
type: RuntimeDefault

dashboard.securityContext -- The container securityContext for the dashboard container

securityContext:
readOnlyRootFilesystem: true
allowPrivilegeEscalation: false
runAsNonRoot: true
runAsUser: 10324
capabilities:
drop:
- ALL
nodeSelector: {}
tolerations: []
affinity: {}

dashboard.topologySpreadConstraints -- Topology spread constraints for the dashboard pods

topologySpreadConstraints: []

Version

8.0.1

Search

• I did search for other open and closed issues before opening this.

Code of Conduct

• I agree to follow this project's Code of Conduct

Additional context

even helloworld pod initializing slower