components
DOK is a deployment tool for Kubernetes clusters, while the coupling degree with Kubernetes itself is not high. As long as the cluster is created through DOK, the management and operation and maintenance of subsequent clusters can be completely separated from DOK. Of course, DOK also integrates many commands and tools related to cluster operation and maintenance, including adding and removing nodes, ETCD backup and recovery, etc. The following is a summary of some configuration information after the DOK cluster is created, and information related to the operation and maintenance of each component. Please check the documentation under the component directory separately.
| components | version | type | remark |
|---|---|---|---|
| OS-release | centos7 | others | |
| Kernel | 3.10.0-1160.71.1.el7.x86_64 | rpm | |
| Kubernetes | v1.24.8 | bin | |
| Containerd | v1.5.5 | bin | |
| runc | v1.1.3 | bin | |
| octant | 0.25.1 | bin | |
| k9s | 0.25.18 | bin | |
| nerdctl | 0.21.0 | bin | |
| helm | 3.6.0 | bin | |
| n9e | 0.0.1 | helm | |
| Flannel | v0.18.1 | yaml | |
| Calico | v3.24.0 | yaml | optional |
| kube-ovn | v1.10.0 | shell | optional |
| multus | v3.9.2 | yaml | optional |
| zsh | 5.0.2 | rpm | optional |
| git | 1.8.3.1 | rpm | optional |
| onmyzsh | master | zip | optional |
| kubectx | 0.9.4 | bin | |
| kubens | 0.9.4 | bin | |
| kubecm | 0.17.0 | bin | |
| krew | 0.17.0 | bin | |
| buildkit | v0.10.3 | bin | |
| knd | master | shell | |
| bat | 0.7.0 | rpm | |
| fzf | master | zip |
# tree -d
.
├── bin [bin related]
│ ├── cni
│ ├── containerd
│ ├── k8s
│ ├── kernel [rpm package containing the kernel and some software]
│ ├── runc
│ └── tools [including nerdctl/k9s/octant/ctr, etc.]
├── conf [config files]
├── docs [Installation and maintenance documents]
├── image [tar]
└── network [CNI plugins]
The premise is that the host has mounted the /data directory. When DOK is installed, it is generally recommended that the host mount the /data directory to a disk with a large space and isolate it from the root directory.
| component | new path | original path |
|---|---|---|
| containerd | /data/containerd/var/lib/containerd/ |
/var/lib/containerd |
The default log address is at /tmp/dok.log
The following deployment results for a cluster of 3 control planes and 3 workers.
# kubectl get pod -A
NAMESPACE NAME READY STATUS RESTARTS AGE
kube-system coredns-558bd4d5db-nvzxq 1/1 Running 0 114m
kube-system coredns-558bd4d5db-ww6cj 1/1 Running 0 114m
kube-system etcd-vm-0-142-centos 1/1 Running 0 113m
kube-system etcd-vm-0-234-centos 1/1 Running 0 111m
kube-system etcd-vm-1-63-centos 1/1 Running 0 114m
kube-system kube-apiserver-vm-0-142-centos 1/1 Running 0 114m
kube-system kube-apiserver-vm-0-234-centos 1/1 Running 1 112m
kube-system kube-apiserver-vm-1-63-centos 1/1 Running 0 114m
kube-system kube-controller-manager-vm-0-142-centos 1/1 Running 0 114m
kube-system kube-controller-manager-vm-0-234-centos 1/1 Running 0 112m
kube-system kube-controller-manager-vm-1-63-centos 1/1 Running 1 114m
kube-system kube-flannel-ds-249s7 1/1 Running 0 112m
kube-system kube-flannel-ds-89fzn 1/1 Running 0 112m
kube-system kube-flannel-ds-bkkr2 1/1 Running 0 112m
kube-system kube-flannel-ds-lx6bn 1/1 Running 0 112m
kube-system kube-flannel-ds-ntkmx 1/1 Running 0 112m
kube-system kube-flannel-ds-xz5bj 1/1 Running 0 112m
kube-system kube-proxy-glc9d 1/1 Running 0 114m
kube-system kube-proxy-hj6g7 1/1 Running 0 112m
kube-system kube-proxy-l4cdq 1/1 Running 0 113m
kube-system kube-proxy-lqq9x 1/1 Running 0 114m
kube-system kube-proxy-s44fg 1/1 Running 0 112m
kube-system kube-proxy-vgwcm 1/1 Running 0 112m
kube-system kube-scheduler-vm-0-142-centos 1/1 Running 0 114m
kube-system kube-scheduler-vm-0-234-centos 1/1 Running 0 112m
kube-system kube-scheduler-vm-1-63-centos 1/1 Running 1 114m