Pod "my-release-fn-fnlb-depl-8f8fcf476-4pvmh" is under "Back-off restarting failed container"

[chatchai-komrangded]

Hi, I tried to deploy fn sever on Kubernetest on OKE, but still facing the issue about "Back-off restarting failed container" Here's my Yaml file..

Default values for Fn

imagePullPolicy: Always

If non-null, will override the cluster name

nameOverride: ''

fn: service: port: 80 type: LoadBalancer annotations: {}

fnlb: image: fnproject/fnlb tag: 0.0.268 logLevel: info resources: {}

fnserver: image: fnproject/fnserver tag: 0.3.461 logLevel: info resources: {} nodeSelector: {} tolerations: []

ui: enabled: true fnui: image: fnproject/ui tag: 0.0.26 resources: {} flowui: image: fnproject/flow tag: ui resources: {} service: flowuiPort: 3000 fnuiPort: 80 type: LoadBalancer annotations: {}

flow: image: fnproject/flow tag: 0.1.83 logLevel: info service: port: 80 type: ClusterIP annotations: {} resources: {}

MySQL chart configuration

mysql: persistence: enabled: false nodeSelector: mysql-storage

If defined, volume.beta.kubernetes.io/storage-class:

## Default: volume.alpha.kubernetes.io/storage-class: default
##
# storageClass:
storageClass: mysql
accessMode: ReadWriteOnce
size: 8Gi

mysqlDatabase: fndb mysqlUser: fnapp mysqlPassword: boomsauce

Redis chart configuration

redis: persistence: enabled: false nodeSelector: redis-storage storageClass: redis accessMode: ReadWriteOnce size: 8Gi usePassword: false

Ingress configuration.

ref: https://kubernetes.io/docs/user-guide/ingress/

ingress: enabled: false

rbac: enabled: false

So Is there any know issue on this?

[chatchai-komrangded]

Here's Log from "my-release-fn-fnlb-depl-8f8fcf476-4pvmh" Pod below:

panic: unknown (get pods) goroutine 26 [running]: github.com/fnproject/fn/fnlb/lb.(*k8sStore).watch(0xc420300ea0, 0xc4202ac220, 0x7, 0x7ffe4b73a3b1, 0x21, 0x50) /go/src/github.com/fnproject/fn/fnlb/lb/k8s_store.go:144 +0xb5e created by github.com/fnproject/fn/fnlb/lb.NewK8sStore /go/src/github.com/fnproject/fn/fnlb/lb/k8s_store.go:56 +0x20e

[vjroxvijay1]

Even I am facing the exact same issue today.

[zootalures]

Can you try with:

rbac:
  enabled: true

in values.yaml?

[zootalures]

Appologies, both - i think this was caused by RBAC being disalbed by default in the helm chart (https://github.com/fnproject/fn-helm/pull/32) and the error message when the LB scans for Fn pods being unclear : https://github.com/fnproject/lb/pull/13

[vjroxvijay1]

Thank you for helping us resolve the issue. That error seems to be gone now. However, I see the following error in load balancer logs.

time="2018-06-26T19:38:32Z" level=error msg="Unable to check API version" error="Get http://10.42.2.43:80/version: dial tcp 10.42.2.43:80: getsockopt: connection refused" node="10.42.2.43:80"

I see all other PODS are up and running successfully

[vjroxvijay1]

It looks like all three fn-api's in the cluster is refusing to connect at :80 with fn load balancer.

[vjroxvijay1]

This issue can be closed as eventually the load balancer determined the nodes are healthy.

[zootalures]

Thanks - and thanks for helping fix the issue!