Feeling Dumb?

[Joshua Skains]

I can't get flannel to act correctly. I have two masters and one node for now just to get to running. But each time a new node or master is added, the flannel-ds container crashes. This is on CentOS 7. Any thoughts?

Thanks!

kube-system   kube-apiserver-kube-master1              1/1       Running             0          52m

kube-system   kube-apiserver-kube-master1-2            1/1       Running             1          45m

kube-system   kube-controller-manager-kube-master1     1/1       Running             0          53m

kube-system   kube-controller-manager-kube-master1-2   1/1       Running             0          46m

kube-system   kube-dns-6f4fd4bdf-8h2zd                 3/3       Running             0          53m

kube-system   kube-flannel-ds-lt768                    0/1       RunContainerError   13         44m

kube-system   kube-flannel-ds-r6fkd                    0/1       CrashLoopBackOff    13         25m

kube-system   kube-flannel-ds-rbp8c                    1/1       Running             0          44m

kube-system   kube-proxy-9csjd                         1/1       Running             0          53m

kube-system   kube-proxy-cpkqv                         1/1       Running             0          47m

kube-system   kube-proxy-j7nxf                         1/1       Running             1          25m

kube-system   kube-scheduler-kube-master1              1/1       Running             0          52m

kube-system   kube-scheduler-kube-master1-2            1/1       Running             0          46m

[Joshua Skains]

BTW, the major error I get is:

Error syncing pod a1576b84-013b-11e8-8a74-005056a112a9 ("kube-flannel-ds-r6fkd_kube-system(a1576b84-013b-11e8-8a74-005056a112a9)"), skipping: failed to "StartContainer" for "kube-flannel" with CrashLoopBackOff: "Back-off 5m0s restarting failed container=kube-flannel pod=kube-flannel-ds-r6fkd_kube-system(a1576b84-013b-11e8-8a74-005056a112a9)"

[Joshua Skains]

So the failed flannel-ds pushes out:

But this doesn't happen on the first flannel-ds

I0124 23:17:06.574471       1 main.go:474] Determining IP address of default interface

I0124 23:17:06.575132       1 main.go:487] Using interface with name eno16777984 and address 10.2.0.176

I0124 23:17:06.575154       1 main.go:504] Defaulting external address to interface address (10.2.0.176)

I0124 23:17:06.595775       1 kube.go:130] Waiting 10m0s for node controller to sync

I0124 23:17:06.595833       1 kube.go:283] Starting kube subnet manager

I0124 23:17:07.596062       1 kube.go:137] Node controller sync successful

I0124 23:17:07.596143       1 main.go:234] Created subnet manager: Kubernetes Subnet Manager - kube-master1-2

I0124 23:17:07.596163       1 main.go:237] Installing signal handlers

I0124 23:17:07.596531       1 main.go:352] Found network config - Backend type: vxlan

I0124 23:17:07.596656       1 vxlan.go:119] VXLAN config: VNI=1 Port=0 GBP=false DirectRouting=false

E0124 23:17:07.597222       1 main.go:279] Error registering network: failed to acquire lease: node "kube-master1-2" pod cidr not assigned

I0124 23:17:07.597552       1 main.go:332] Stopping shutdownHandler...

[Joshua Skains]

I got further. It didn't like the 192.168.0.0/24 network I gave it. But now it locks making the DNS service:

Jan 24 17:01:45 [localhost] kubelet: E0124 17:01:45.166571    9221 pod_workers.go:186] Error syncing pod 51a70790-0162-11e8-a50b-005056a112a9 ("kube-dns-6f4fd4bdf-568pl_kube-system(51a70790-0162-11e8-a50b-005056a112a9)"), skipping: failed to "CreatePodSandbox" for "kube-dns-6f4fd4bdf-568pl_kube-system(51a70790-0162-11e8-a50b-005056a112a9)" with CreatePodSandboxError: "CreatePodSandbox for pod \"kube-dns-6f4fd4bdf-568pl_kube-system(51a70790-0162-11e8-a50b-005056a112a9)\" failed: rpc error: code = Unknown desc = NetworkPlugin cni failed to set up pod \"kube-dns-6f4fd4bdf-568pl_kube-system\" network: failed to set bridge addr: \"cni0\" already has an IP address different from 172.27.0.1/24"