[Kubernetes] k8s multi cluster switching [including detailed explanation of context function] [including handling method of 6443 port unreachable]

Posted by mj99 on Fri, 24 Dec 2021 14:01:58 +0100

What is multi cluster?

  • If you don't create a cluster, first install k8s and build a cluster. One of my blog categories is: "kubernetes" k8s installation details [create a cluster, join a cluster, kick out a cluster, reset a cluster...]. You can have a look at it if you want to know.

  • As shown in the following figure, a cluster has a master node and a node node
    They have nothing to do with each other.
    As follows, I created two clusters and started them all.

  • Each cluster provides services independently, and the two are not related to each other. Generally, we manage multiple clusters by logging in to the master node of a cluster through ssh
    After multi clustering, you can manage another master node on a master node without ssh login.

  • The configurations of the two clusters are exactly the same, except that the ip addresses are different. I will explain these two nodes below
    The host name and ip address are as follows

#Cluster 1
[root@master ~]# kubectl get nodes
NAME     STATUS   ROLES                  AGE    VERSION
master   Ready    control-plane,master   5d2h   v1.21.0
node1    Ready    <none>                 5d2h   v1.21.0
node2    Ready    <none>                 5d1h   v1.21.0
[root@master ~]# 
[root@master ~]# ip a | grep 59
    inet 192.168.59.142/24 brd 192.168.59.255 scope global noprefixroute dynamic ens33
[root@master ~]# 
[root@master ~]# hostname
master
[root@master ~]# 

#Cluster 2
[root@master2 ~]# kubectl get node
NAME     STATUS   ROLES                  AGE    VERSION
master   Ready    control-plane,master   5d2h   v1.21.0
node1    Ready    <none>                 5d2h   v1.21.0
node2    Ready    <none>                 5d1h   v1.21.0
[root@master2 ~]# 
[root@master2 ~]# ip a | grep 59
    inet 192.168.59.151/24 brd 192.168.59.255 scope global noprefixroute ens33
[root@master2 ~]# 
[root@master2 ~]# hostname
master2
[root@master2 ~]#

kubeconfig file content preparation [master node operation]

  • The default kubeconfig file is: ~/ kube/config
  • View the kubeconfig file command: kubectl config view
[root@master .kube]# kubectl config view 
apiVersion: v1
clusters:
- cluster:
    certificate-authority-data: DATA+OMITTED
    server: https://192.168.59.142:6443
  name: kubernetes
contexts:
- context:
    cluster: kubernetes
    namespace: default
    user: kubernetes-admin
  name: kubernetes-admin@kubernetes
current-context: kubernetes-admin@kubernetes
kind: Config
preferences: {}
users:
- name: kubernetes-admin
  user:
    client-certificate-data: REDACTED
    client-key-data: REDACTED
[root@master .kube]#

Single cluster configuration file modification

  • You want to modify the information of the current cluster.
  • The configuration file is ~/ kube/config. See mode 1 below for the code explanation. The modification method is the same as mode 1, except for leaving a cluster information.

Method 1: manual editing

File backup

  • This is the most original method. It is implemented directly by editing the configuration file. When doing this, be sure to back up the file [key is here].
[root@master ~]# cd .kube/
[root@master .kube]# cp config config.bak
[root@master .kube]#

Code explanation of config file [detailed explanation of context]

  • After the backup, we will edit it directly: ~/ kube/config file
    In order to facilitate editing, it is recommended to delete all three keys after entering to get the following refreshing file content [if not deleted, too many keys will affect the operation]
    In order to facilitate understanding, I directly add comments after the following code. After careful reading, the function of each line of code is clear [the following is the original code of the file, without any modification, only for explanation]
  1 apiVersion: v1
  2 clusters: # This means that the cluster to which you need to connect [lines 3-6 below are a group] and how many new clusters you need to add, copy lines 3-6, add them to line 6, and modify the relevant information
  3 - cluster: # Cluster start [3-6 rows in a group]
  4     certificate-authority-data: #This is the key. The viewing method will be described later
  5     server: https://192.168. 59.142:6443 # the ip address of the master node of this cluster
  6   name: kubernetes #Cluster name, which can be customized
  7 contexts:#Context, bind the upper cluster [clusters] and the lower user [users] together [they are independent before binding]
  8 - context: #At the beginning of the context [8-12 acts as a group], how many clusters and how many contexts are added are placed under line 12
  9     cluster: kubernetes #This is the name of the cluster above
 10     namespace: default #Namespace [can be modified]
 11     user: kubernetes-admin #This is the name of the following user
 12   name: kubernetes-admin@kubernetes # This context name can be customized
 13 current-context: kubernetes-admin@kubernetes # This is the default context name [cluster name] [for example, if there are multiple clusters, the cluster will be used by default]
 14 kind: Config
 15 preferences: {}
 16 users: # This specifies the user information [17-20 below as a group], the number of clusters and the number of users added [copy content below 20 lines]
 17 - name: kubernetes-admin  # User name can be customized
 18   user:
 19     client-certificate-data: #This is the key. The viewing method will be described later
 20     client-key-data:#This is the key. The viewing method will be described later

config code editing

According to the above explanation, make relevant edits. Here is the code I configured

[root@master .kube]# vim config
  1 apiVersion: v1
  2 clusters:
  3 - cluster:
  4     certificate-authority-data:
  5     server: https://192.168.59.142:6443
  6   name: master
  7 - cluster:
  8     certificate-authority-data:
  9     server: https://192.168.59.151:6443
 10   name: master1
 11 contexts:
 12 - context:
 13     cluster: master
 14     namespace: default
 15     user: ccx
 16   name: context
 17 - context:
 18     cluster: master1
 19     namespace: default
 20     user: ccx1
 21   name: context1
 22 current-context: context
 23 kind: Config
 24 preferences: {}
 25 users:
 26 - name: ccx
 27   user:
 28     client-certificate-data:
 29     client-key-data:
 30 - name: ccx1
 31   user:
 32     client-certificate-data:
 33     client-key-data:

View certificate [key]

  • Because the key of the current cluster has been deleted above, you need to copy it back and copy the keys of other clusters.
    The key s of each cluster are different. Do not copy the cross.

  • key file: ~ / kube/config [yes, the file edited above]

    • For the edited cluster, view the config file backed up at the beginning [reopen a window]. For other cluster information, directly view the corresponding copy of the config file.
    • I don't need to say more about the corresponding replication. If there are multiple clusters, which two are the specified clusters and users in the context, just don't make mistakes when copying key s.
    • Note: there is a space in front of the key, and there is no space at the end. Don't have more or less spaces.

  • After all, the code is as follows:

[root@master .kube]# cat config
apiVersion: v1
clusters:
- cluster:
    certificate-authority-data: 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
    server: https://192.168.59.142:6443
  name: master
- cluster:
    certificate-authority-data: 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
    server: https://192.168.59.151:6443
  name: master1
contexts:
- context:
    cluster: master
    namespace: default
    user: ccx
  name: context
- context:
    cluster: master1
    namespace: default
    user: ccx1
  name: context1
current-context: context
kind: Config
preferences: {}
users:
- name: ccx
  user:
    client-certificate-data: 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
    client-key-data: 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
- name: ccx1
  user:
    client-certificate-data: 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
    client-key-data: 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
[root@master .kube]#

At this point, the multi cluster configuration is complete.

Processing of unreachable error messages on cluster 6443 port

  • The error message is that the IP port of your current cluster is unreachable
The connection to the server 192.168.59.151:6443 was refused - did you specify the right host or port?
  • Idea:
    • 1. Check whether the docker service is active: systemctl is active docker
    • 2. Check whether the kubectl service is active: systemctl is active kubelet
    • 3. Check configuration file: ~ / Is kube/config incorrect
    • 4. Check whether the image is missing: docker images [if the image is missing, supplement it, and then restart the above two services, all images will be started automatically (do not try to start the image manually, restart the above two services in order, and docker will automatically pull up the services in order)]
    • 5. Check whether the port is listening: netstat -ntlp | grep 6443
    • 6. Check whether the service is normal: docker ps [there should be about 17 normal]
    • 7. Check the system log: journalctl -xeu kubelet
[root@master ~]# systemctl is-active docker
active
[root@master ~]# systemctl is-active kubelet
active
[root@master ~]# 
[root@master ~]# netstat -ntlp | grep 6443
tcp6       0      0 :::6443                 :::*                    LISTEN      19672/kube-apiserve 
[root@master ~]# 
[root@master ~]# docker ps | wc -l
17
[root@master ~]#
  • But I think this is special. All the services of my problem cluster are normal and can be used alone in the problem cluster. All kinds of troubleshooting are fruitless. Finally, I thought that my second cluster was cloned. I did find that the key s of my two clusters are the same through the configuration file. Therefore, theoretically, the two clusters are the same, and the IP cannot be changed.
    Now the error is reported as follows
[root@master ~]# kubectl config use-context context1
Switched to context "context1".
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME       CLUSTER   AUTHINFO   NAMESPACE
          context    master    ccx        default
*         context1   master1   ccx1       default
[root@master ~]# 
[root@master ~]# kubectl get nodes
The connection to the server 192.168.59.151:6443 was refused - did you specify the right host or port?
[root@master ~]#
  • In order to verify my idea, I changed the ip of the second cluster in the configuration file to be the same as that of my first cluster, and did not modify anything else [the key is still the key of the second cluster], and then accessed it again. It was successful
[root@master ~]# cat .kube/config | grep server
    server: https://192.168.59.142:6443
    server: https://192.168.59.142:6443
[root@master ~]# 
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME       CLUSTER   AUTHINFO   NAMESPACE
          context    master    ccx        default
*         context1   master1   ccx1       default
[root@master ~]# 
[root@master ~]# kubectl get nodes
NAME     STATUS   ROLES                  AGE     VERSION
master   Ready    control-plane,master   6d      v1.21.0
node1    Ready    <none>                 6d      v1.21.0
node2    Ready    <none>                 5d22h   v1.21.0
[root@master ~]# 
[root@master ~]# kubectl get pods
No resources found in default namespace.
[root@master ~]# kubectl get pods -n ccx
NAME                          READY   STATUS             RESTARTS   AGE
centos-7846bf67c6-s9tmg       0/1     ImagePullBackOff   0          23h
nginx-test-795d659f45-j9m9b   0/1     ImagePullBackOff   0          24h
nginx-test-795d659f45-txf8l   0/1     ImagePullBackOff   0          24h
[root@master ~]#
  • The conclusion is: master cluster cannot be cloned!!!!!!! The pit summarized in most of the day
    Of course, cloning is not impossible. One article in my blog is: "[kubernetes] k8s installation details [create cluster, join cluster, kick cluster, reset cluster...]", just reset the cluster according to the methods in this, and then repeat the above steps, and everything will be normal
    Tearful eyes... It took most of the day to sum up the pit
[root@master ~]# cat .kube/config| grep server
    server: https://192.168.59.142:6443
    server: https://192.168.59.151:6443
[root@master ~]# 
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME           CLUSTER   AUTHINFO   NAMESPACE
*         context        master    ccx        default
          context1-new   master1   ccx1       default
[root@master ~]# kubecon^C
[root@master ~]# 
[root@master ~]# kubectl config use-context context1-new 
Switched to context "context1-new".
[root@master ~]# kubectl get pods
No resources found in default namespace.
[root@master ~]# kubectl get pods -n ccx
No resources found in ccx namespace.
[root@master ~]# kubectl get pods -n kube-system 
NAME                                       READY   STATUS    RESTARTS   AGE
calico-kube-controllers-78d6f96c7b-q25lf   1/1     Running   0          7m36s
calico-node-g2wfj                          1/1     Running   0          7m36s
calico-node-gtfbc                          1/1     Running   0          7m36s
calico-node-nd29t                          1/1     Running   0          6m49s
coredns-545d6fc579-j4ckc                   1/1     Running   0          12m
coredns-545d6fc579-qn4kz                   1/1     Running   0          12m
etcd-master2                               1/1     Running   0          12m
kube-apiserver-master2                     1/1     Running   1          12m
kube-controller-manager-master2            1/1     Running   0          12m
kube-proxy-c4x8j                           1/1     Running   0          12m
kube-proxy-khh2z                           1/1     Running   0          10m
kube-proxy-pckf7                           1/1     Running   0          6m49s
kube-scheduler-master2                     1/1     Running   0          12m
[root@master ~]# 
[root@master ~]#

kubectl get node no node

  • Other services are normal [see services above], that is, no node is found or missing on the master, as shown below
[root@master ~]# kubectl  get nodes
NAME     STATUS   ROLES    AGE     VERSION
master   Ready    <none>   16m     v1.21.0
node1    Ready    <none>   2m19s   v1.21.0
[root@master ~]#
  • resolvent
    Log in to the undiscovered node node, restart the docker and kubelet services, and return to the master node later to see the node information
[root@master ~]# kubectl  get nodes
NAME     STATUS   ROLES    AGE     VERSION
master   Ready    <none>   16m     v1.21.0
node1    Ready    <none>   2m19s   v1.21.0
[root@master ~]# 
[root@master ~]# ssh node2
root@node2's password: 
Last login: Fri Jul  2 12:19:44 2021 from 192.168.59.1
[root@node2 ~]# systemctl restart docker
[root@node2 ~]# systemctl restart kubelet
[root@node2 ~]# 
[root@node2 ~]# exit
logout
Connection to node2 closed.
[root@master ~]# 
[root@master ~]# kubectl  get nodes
NAME     STATUS     ROLES    AGE    VERSION
master   Ready      <none>   20m    v1.21.0
node1    Ready      <none>   7m5s   v1.21.0
node2    NotReady   <none>   2s     v1.21.0
[root@master ~]#

Method 2: command generation

config multi cluster management (contexts)

View all cluster information

Command: kubectl config get contexts

[root@master ~]# kubectl config get-contexts 
CURRENT   NAME       CLUSTER   AUTHINFO   NAMESPACE
          context    master    ccx        default
*         context1   master1   ccx1       default
[root@master ~]#
  • Command interpretation
    • CURRENT: * indicates which cluster you are currently in
    • NAME: context NAME
    • CLUSTER: CLUSTER name
    • AUTHINFO: user name
    • NAMESPACE: NAMESPACE

Context switching [cluster switching]

  • Command: kubectl config use context context NAME [NAME in kubectl config get contexts is the context NAME]
  • For example, I switch to context
    *The location is the current cluster.
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME       CLUSTER   AUTHINFO   NAMESPACE
          context    master    ccx        default
*         context1   master1   ccx1       default
[root@master ~]# 
[root@master ~]# kubectl config use-context context
Switched to context "context".
[root@master ~]# 
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME       CLUSTER   AUTHINFO   NAMESPACE
*         context    master    ccx        default
          context1   master1   ccx1       default
[root@master ~]#

Rename context

  • Command: kubectl config rename context current name new name
  • For example, I modified context1 to context1 new
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME       CLUSTER   AUTHINFO   NAMESPACE
*         context    master    ccx        default
          context1   master1   ccx1       default
[root@master ~]# 
[root@master ~]# kubectl config rename-context context1 context1-new
Context "context1" renamed to "context1-new".
[root@master ~]# 
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME           CLUSTER   AUTHINFO   NAMESPACE
*         context        master    ccx        default
          context1-new   master1   ccx1       default
[root@master ~]#

Create a context [not recommended]

  • Command: kubectl config set context custom context name
  • It is meaningless to simply create such a cluster, because there is no designated cluster, and there is no designated user. [it needs to be used in combination with the following modification information]
    For example, when I create a my context, I can see nothing
[root@master ~]# kubectl config set-context my-context
Context "my-context" created.
[root@master ~]#
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME         CLUSTER   AUTHINFO   NAMESPACE
          context      master    ccx        default
*         context1     master1   ccx1       default
          my-context                        
[root@master ~]#
  • As above, after creation, it will be in: ~ / kube/config generates a context, but there is no other key information. You need to manually specify ns, user and name by modifying the information
- context:
    cluster: my-context
    namespace: 
    user: 
  name:

Modify information

  • This is equivalent to two functions
  • 1. Modify the context information that has been created
  • 2. Add context content [add remaining information after creating with command]
  • Syntax:
kubectl  config set-context my-context[Existing context name]  --namespace=ccx[Must be an existing namespace] --cluster=my-context[User defined name (modification only)] --user=my-user[User defined name (modification only)]
  • Non custom view commands
    • Context view: kubectl config get contexts
    • Namespace view: kubectl get ns

Special instructions for error reporting of new information

  • As I said earlier, modifying information is not recommended to add context information because:
    A cluster has only one set of key s, and theoretically only one context can exist. If the context is created by command and other information is customized, in ~ / Only the following information will be generated in kube/config file, but neither cluster nor user exists, so an error will be reported when using this file, indicating that there is no 8080 port. The reason is that this context does not specify the server and user name, so the cluster information cannot be found [in this case, you can bind cluster and user to the existing cluster]
- context:
    cluster: my-context
    namespace: ccx
    user: my-user
  name: my-context
  • For example, I added information to my context created by the above command
    After I switch to this context, I am prompted that port 8080 is not started [this is the error message, which is an example of an error, so do not create a context with a command, and specify it directly by modifying the configuration file]
[root@master ~]# kubectl  config set-context my-context --namespace=ccx --cluster=my-context --user=my-user
Context "my-context" modified.
[root@master ~]#
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME         CLUSTER      AUTHINFO   NAMESPACE
*         context      master       ccx        default
          context1     master1      ccx1       default
          my-context   my-context   my-user    ccx
[root@master ~]# 
[root@master ~]# kubectl config use-context my-context 
Switched to context "my-context".
[root@master ~]# 
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME         CLUSTER      AUTHINFO   NAMESPACE
          context      master       ccx        default
          context1     master1      ccx1       default
*         my-context   my-context   my-user    ccx
[root@master ~]# 
[root@master ~]# kubectl get pods
The connection to the server localhost:8080 was refused - did you specify the right host or port?
[root@master ~]# kubectl get nodes
The connection to the server localhost:8080 was refused - did you specify the right host or port?
[root@master ~]# 
[root@master ~]# netstat -ntlp | grep 8080
[root@master ~]#

Delete context

  • Command: kubectl config delete context context name
    If you don't know what you're doing, don't delete it. [generally, a cluster has only one context, which cannot be used after deleting the current cluster [generally used to delete multiple clusters]]
    After deleting the context through the command, the configuration file ~ / Relevant contents in kube/config will also be deleted
  • For example, I delete the name my context [this is the context I created through the command above and can't be used, so I can kill it directly]
[root@master ~]# 
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME         CLUSTER      AUTHINFO   NAMESPACE
*         context      master       ccx        default
          context1     master1      ccx1       default
          my-context   my-context   my-user    ccx
[root@master ~]# kubectl config delete-context my-context 
deleted context my-context from /root/.kube/config
[root@master ~]# 
[root@master ~]# kubectl config get-contexts 
CURRENT   NAME       CLUSTER   AUTHINFO   NAMESPACE
*         context    master    ccx        default
          context1   master1   ccx1       default
[root@master ~]# 
[root@master ~]# cat ~/.kube/config | grep my-con
[root@master ~]#

Topics: Kubernetes

Hot Topics

©2024 Programmer Think