K
Q

OKD 3.11 Installation failed "Control plane pods didn't come up" "network plugin is not ready: cni config uninitialized"

4/18/2019

OKD 3.11 Installation failed "Control plane pods didn't come up"

Environment

  • CentOS Linux release 7.6.1810 (Core)
  • ansible 2.6.16
  • OKD 3.11
  • Docker version 1.13.1, build b2f74b2/1.13.1

Ansible inventory file

    ##Open shift master nodes
# Create an OSEv3 group that contains the masters, nodes, and etcd groups
[OSEv3:children]
masters
nodes
etcd
# host group for masters
[masters]
SBSTJVMLX605 openshift_ip=192.168.62.95
# host group for etcd
[etcd]
SBSTJVMLX605 openshift_ip=192.168.62.95
# host group for nodes, includes region info
[nodes]
SBSTJVMLX605 openshift_node_group_name='node-config-master-infra' openshift_schedulable=true openshift_ip=192.168.62.95
[OSEv3:vars]
openshift_deployment_type=origin
#openshift_release="3.11"
openshift_disable_check=disk_availability,docker_storage,package_version,memory_availability
openshift_enable_docker_excluder=false
debug_level=4

Errors from ansible

Failure summary:
  1. Hosts:    SBSTJVMLX605
     Play:     Configure masters
     Task:     Report control plane errors
     Message:  Control plane pods didn't come up

Errors from journal ctl

31207 kubelet.go:2101] Container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:docker: network plugin is not ready: cni config uninitialized

Also

origin-node[31207]: W0418 10:26:47.866972   31207 docker_service.go:545] Hairpin mode set to "promiscuous-bridge" but kubenet is not enabled, falling back to "hairpin-veth"
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.866991   31207 docker_service.go:238] Hairpin mode set to "hairpin-veth"
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: W0418 10:26:47.867101   31207 cni.go:172] Unable to update cni config: No networks found in /etc/cni/net.d
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: W0418 10:26:47.870697   31207 cni.go:172] Unable to update cni config: No networks found in /etc/cni/net.d
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.870720   31207 plugins.go:159] Loaded network plugin "cni"
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.870745   31207 docker_service.go:253] Docker cri networking managed by cni
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.883666   31207 docker_service.go:258] Docker Info: &{ID:BHX3:QGBT:WAEZ:TNJL:PCFX:4S6K:CEV6:QMMQ:NELX:RCXS:EXOE:AXYY Containers:0 ContainersRunning:0 ContainersPaused:0 ContainersStopped:0 Images:5 Driver:overlay2 DriverStatus:[[Backing Filesystem xfs] [Supports d_type true] [Native Overlay Diff true]] SystemStatus:[] Plugins:{Volume:[local] Network:[bridge host macvlan null overlay] Authorization:[] Log:[]} MemoryLimit:true SwapLimit:true KernelMemory:true CPUCfsPeriod:true CPUCfsQuota:true CPUShares:true CPUSet:true IPv4Forwarding:true BridgeNfIptables:false BridgeNfIP6tables:false Debug:false NFd:16 OomKillDisable:true NGoroutines:23 SystemTime:2019-04-18T10:26:47.873639519+05:30 LoggingDriver:json-file CgroupDriver:systemd NEventsListener:0 KernelVersion:3.10.0-957.el7.x86_64 OperatingSystem:CentOS Linux 7 (Core) OSType:linux Architecture:x86_64 IndexServerAddress:https://index.docker.io/v1/ RegistryConfig:0xc4208af5e0 NCPU:4 MemTotal:8370298880 GenericResources:[] DockerRootDir:/var/lib/docker HTTPProxy:http://sbstjsvfw01:8080/ HTTPSProxy:http://sbstjsvfw01:8080/ NoProxy: 192.168.62.95, localhost,.cluster.local,.svc,127.0.0.1,169.254.169.254,172.30.0.1,192.168.62.95,sbstjvmlx605.suntecsbs.com Name:SBSTJVMLX605.SUNTECSBS.COM Labels:[] ExperimentalBuild:false ServerVersion:1.13.1 ClusterStore: ClusterAdvertise: Runtimes:map[docker-runc:{Path:/usr/libexec/docker/docker-runc-current Args:[]} runc:{Path:docker-runc Args:[]}] DefaultRuntime:docker-runc Swarm:{NodeID: NodeAddr: LocalNodeState:inactive ControlAvailable:false Error: RemoteManagers:[] Nodes:0 Managers:0 Cluster:0xc420c46dc0} LiveRestoreEnabled:false Isolation: InitBinary:/usr/libexec/docker/docker-init-current ContainerdCommit:{ID: Expected:aa8187dbd3b7ad67d8e5e3a15115d3eef43a7ed1} RuncCommit:{ID:df5c38a9167e87f53a9894d77c0950e178a745e7 Expected:9df8b306d01f59d3a8029be411de015b7304dd8f} InitCommit:{ID:fec3683b971d9c3ef73f284f176672c44b448662 Expected:949e6facb77383876aeff8a6944dde66b3089574} SecurityOptions:[name=seccomp,profile=/etc/docker/seccomp.json name=selinux]}
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.883740   31207 docker_service.go:271] Setting cgroupDriver to systemd
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.883810   31207 kubelet.go:628] Starting the GRPC server for the docker CRI shim.
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.883872   31207 docker_server.go:59] Start dockershim grpc server
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: W0418 10:26:47.883909   31207 util_unix.go:75] Using "/var/run/dockershim.sock" as endpoint is deprecated, please consider using full url format "unix:///var/run/dockershim.sock".
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.890481   31207 oom_linux.go:65] attempting to set "/proc/19601/oom_score_adj" to "-999"
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.890657   31207 oom_linux.go:65] attempting to set "/proc/19610/oom_score_adj" to "-999"
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.900614   31207 remote_runtime.go:43] Connecting to runtime service /var/run/dockershim.sock
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: W0418 10:26:47.900638   31207 util_unix.go:75] Using "/var/run/dockershim.sock" as endpoint is deprecated, please consider using full url format "unix:///var/run/dockershim.sock".
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.900715   31207 remote_image.go:40] Connecting to image service /var/run/dockershim.sock
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: W0418 10:26:47.900728   31207 util_unix.go:75] Using "/var/run/dockershim.sock" as endpoint is deprecated, please consider using full url format "unix:///var/run/dockershim.sock".
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.900860   31207 plugins.go:56] Registering credential provider: .dockercfg
Apr 18 10:26:47 SBSTJVMLX605.SUNTECSBS.COM origin-node[31207]: I0418 10:26:47.902596   31207 kuberuntime_manager.go:186] Container runtime docker initialized, version: 1.13.1, apiVersion: 1.26.0

Also a lot of connect refused on 8443 port master api.

Did some one came across some thing similar. I have a running cluster on 3.9 and was working properly. I tried this as a fresh installation after uninstalling 3.9

-- arunvg
kubernetes
okd
openshift
openshift-3
openshift-origin

Similar Questions

Logically seperate azure kubernetes deployments
kubectl replace -f creates pod indefinitely in Pending state
how to enable kubelet logging verbosity
How to use helm chart for spark on k8s
Installing helm tiller on running K8s cluster failed to fetch stable helm repository url with error 403 forbidden
How to access a service deployed as NodePort in Minikube in EC2 from outside
Is there a way to limit per_process memory on a shared vgpu with Spacy, Torch, etc?
Find logs of POD in AKS using Log Analytics Query
How can I provision a volume (part II)?
To make a LoadBalancer service internal on GKE
Pod name resolution for StatefulSet doesn't work

2 Answers

4/18/2019

Oof could get the thing up !!

  1. Unable to update cni config: No networks found in /etc/cni/net.d. The network plugin was not ready perhaps because of this. This was the thing @capt2101akash pointed also.

    Got a work around solution from the below mailing list

    Reference solution

    Basically created a file /etc/cni/net.d/80-openshift-network.conf with content

    { "cniVersion": "0.2.0", "name": "openshift-sdn", "type": "openshift-sdn" }

  2. Again it was not coming up, Ansible scripts showed me the same old story "Control planes not up"

    This time master logs /usr/local/bin/master-logs api api was showing a command not found in /etc/origin/master/master.env . Guessed a problem in inventory file. The entry for proxy had an additional \ and no proxy some spaces.

    Santized the inventory file.

    Redeployed and it is up and accessible.

    This was a workaround, would post it if i get into the real cause.

-- arunvg
Source: StackOverflow
4/18/2019

You need a network plugin for your installation to complete. It is failing because it can't find the network plugin to use for ex. CNI, weave-net, etc. Here you can find the listed plugins supported by k8s.

-- capt2101akash
Source: StackOverflow