舊版的 calico 沒有用到 /sys/fs/bpf 可以直接 apply yaml
curl wget https://docs.projectcalico.org/v3.10/manifests/calico.yaml -O
找到 CALICO_IPV4POOL_CIDR 進行修改
加上這組設定 (這是gateway IP)
- name: IP_AUTODETECTION_METHOD
value: can-reach=10.168.100.253
新版可能會出現錯誤
path /sys/fs is mounted on /sys but it is not a shared mount
因為 OpenRC 沒有把 mount 轉換成 shared mount
findmnt -o TARGET,PROPAGATION / 進行確認
可以手動 mount --make-rshared / 或加在開機 script
使用 kubectl -n kube-system get pods -w 檢查 calico 已正常運作
如下
NAME READY STATUS RESTARTS AGE
calico-kube-controllers-744cfdf676-5l8d6 1/1 Running 0 35s
calico-node-4mxc7 1/1 Running 0 35s
calico-node-77wxr 1/1 Running 0 35s
calico-node-8rw75 1/1 Running 0 35s
calico-node-nnc25 1/1 Running 0 35s
calico-node-p27l4 1/1 Running 0 35s
calico-node-tlkrh 1/1 Running 0 35s
coredns-74ff55c5b-spddd 1/1 Running 0 24m
coredns-74ff55c5b-trzpg 1/1 Running 0 24m
kube-apiserver-k8s-node1 1/1 Running 0 24m
kube-apiserver-k8s-node2 1/1 Running 0 22m
kube-apiserver-k8s-node3 1/1 Running 0 21m
kube-controller-manager-k8s-node1 1/1 Running 0 24m
kube-controller-manager-k8s-node2 1/1 Running 0 22m
kube-controller-manager-k8s-node3 1/1 Running 0 21m
kube-proxy-4v5rd 1/1 Running 0 19m
kube-proxy-8wjkc 1/1 Running 0 20m
kube-proxy-brq2p 1/1 Running 0 24m
kube-proxy-jjh5q 1/1 Running 0 21m
kube-proxy-mdph2 1/1 Running 0 18m
kube-proxy-shvh7 1/1 Running 0 22m
kube-scheduler-k8s-node1 1/1 Running 0 24m
kube-scheduler-k8s-node2 1/1 Running 0 22m
kube-scheduler-k8s-node3 1/1 Running 0 21m
ref: https://success.mirantis.com/article/not-a-shared-mount-error
