Ta-Ching ChenandGitHub
bfd005a295
ReplicaSet generated by Deployment distributes pods to nodes based on nodes resource usage, which means in some cases the newly created pods may be assigned to only a few nodes. The requests will go to the same node if there multiple pods on it and may increase the node resource usage and overall request latency. To solve the problem mentioned above, this PR allows users to deploy the router as DaemonSet so that we can distribute requests across all nodes for better workload distribution and lower latency.
153 lines
4.8 KiB
YAML
153 lines
4.8 KiB
YAML
#
|
|
# Fission chart configuration
|
|
#
|
|
|
|
## Kubernetes configuration
|
|
## For minikube, set this to NodePort, elsewhere use LoadBalancer or ClusterIP.
|
|
serviceType: ClusterIP
|
|
|
|
## For minikube, set this to NodePort, elsewhere use LoadBalancer or ClusterIP.
|
|
routerServiceType: LoadBalancer
|
|
|
|
## Image base repository
|
|
## Leave it empty for using existing local image
|
|
repository: index.docker.io
|
|
|
|
## Fission image repository
|
|
image: fission/fission-bundle
|
|
|
|
## Fission image version
|
|
imageTag: 1.5.0
|
|
|
|
## Image pull policy
|
|
pullPolicy: IfNotPresent
|
|
|
|
## Fission fetcher repository
|
|
fetcherImage: fission/fetcher
|
|
|
|
## Fission fetcher image version
|
|
fetcherImageTag: 1.5.0
|
|
|
|
## Port at which Fission controller service should be exposed
|
|
controllerPort: 31313
|
|
|
|
## Port at which Fission router service should be exposed
|
|
routerPort: 31314
|
|
|
|
## Namespace in which to run fission functions (this is different from
|
|
## the release namespace)
|
|
functionNamespace: fission-function
|
|
|
|
## Namespace in which to run fission builders (this is different from
|
|
## the release namespace)
|
|
builderNamespace: fission-builder
|
|
|
|
## Enable istio integration
|
|
enableIstio: false
|
|
|
|
## Router config
|
|
router:
|
|
deployAsDaemonSet: false
|
|
svcAddressMaxRetries: 5
|
|
svcAddressUpdateTimeout: 30s
|
|
## Add annotations for router
|
|
# svcAnnotations:
|
|
# cloud.google.com/load-balancer-type: Internal
|
|
roundTrip:
|
|
## If true, router will disable the HTTP keep-alive which result in performance degradation.
|
|
## But it ensures that router can redirect new coming requests to new function pods.
|
|
##
|
|
## If false, router will enable transport keep-alive feature for better performance.
|
|
## However, the drawback is it takes longer to switch to newly created function pods
|
|
## if using newdeploy as executor type for function. If you want to preserve the
|
|
## performance while keeping the short switching time to new function, you can create
|
|
## an environment with short grace period by setting flag "--graceperiod" (default 360s),
|
|
## so that kubernetes will be able to reap old function pod quickly.
|
|
##
|
|
## For details, see https://github.com/fission/fission/issues/723
|
|
disableKeepAlive: true
|
|
|
|
## The keep-alive period for an active network connection to function pod.
|
|
keepAliveTime: 30s
|
|
|
|
## HTTP transport request timeout
|
|
timeout: 50ms
|
|
|
|
## The length of request timeout will multiply with timeoutExponent after each retry
|
|
timeoutExponent: 2
|
|
|
|
## Max retries times of a failed request
|
|
maxRetries: 10
|
|
|
|
## Persist data to a persistent volume.
|
|
persistence:
|
|
## If true, fission will create/use a Persistent Volume Claim
|
|
## If false, use emptyDir
|
|
##
|
|
enabled: true
|
|
|
|
## A manually managed Persistent Volume Claim name
|
|
## Requires persistence.enabled: true
|
|
## If defined, PVC must be created manually before volume will be bound
|
|
##
|
|
# existingClaim:
|
|
|
|
## If defined, storageClassName: <storageClass>
|
|
## If set to "-", storageClassName: "", which disables dynamic provisioning
|
|
## If undefined (the default) or set to null, no storageClassName spec is
|
|
## set, choosing the default provisioner. (gp2 on AWS, standard on
|
|
## GKE, AWS & OpenStack)
|
|
##
|
|
# storageClass: "-"
|
|
|
|
accessMode: ReadWriteOnce
|
|
size: 8Gi
|
|
|
|
## Extend the container specs for the core fission pods.
|
|
## Can be used to add things like affinty/tolerations/nodeSelectors/etc.
|
|
## For example:
|
|
## extraCoreComponentPodConfig:
|
|
## affinity:
|
|
## nodeAffinity:
|
|
## requiredDuringSchedulingIgnoredDuringExecution:
|
|
## nodeSelectorTerms:
|
|
## - matchExpressions:
|
|
## - key: capability
|
|
## operator: In
|
|
## values:
|
|
## - app
|
|
extraCoreComponentPodConfig:
|
|
# affinity:
|
|
# tolerations:
|
|
# nodeSelector:
|
|
|
|
## Analytics let us count how many people installed fission. Set to
|
|
## false to disable analytics.
|
|
analytics: true
|
|
|
|
## Internally used for generating an analytics job for non-helm installs
|
|
analyticsNonHelmInstall: false
|
|
|
|
## Archive pruner is a garbage collector for archives on the fission storage service.
|
|
## This interval configures the frequency at which it runs inside the storagesvc pod.
|
|
## The value is in minutes.
|
|
pruneInterval: 60
|
|
|
|
## Fission pre-install/pre-upgrade checks live in this image
|
|
preUpgradeChecksImage: fission/pre-upgrade-checks
|
|
|
|
## if there are any pod specialization errors when a function is triggered and this flag is set to true, the error
|
|
## summary is returned as part of http response
|
|
debugEnv: false
|
|
|
|
## set this flag to true if prometheus needs to be deployed along with fission
|
|
prometheusDeploy: false
|
|
|
|
## set this flag to false if you dont need canary deployment feature
|
|
canaryDeployment:
|
|
enabled: false
|
|
|
|
# Use these flags to enable opentracing, the variable is endpoint of Jaeger collector in the format shown below
|
|
#traceCollectorEndpoint: "http://jaeger-collector.jaeger.svc:14268/api/traces?format=jaeger.thrift"
|
|
#traceSamplingRate: 0.75
|