Kubernetes Certificate Expire Causes Cluster Wide Communication Halt

Introduction

This document describes a possible outage issue that the customers might face when they have a Kubernetes based system that has been installed for more than 365 days. Furthermore, it goes through the steps needed to fix the situation and get the Kubernetes based system back up and running.

Problem

After one year of time of default installed Kubernetes cluster, the client certificates expire. You will not be able to access Cisco CloudCentre Suite (CCS). Although it will still appear up, you will not be able to log in. If you navigate to the kubectl CLI, you will see this error, "Unable to connect to the server: x509: certificate has expired or is not yet valid."

You can run this bash script to see the expiration date of their certificates:

for crt in /etc/kubernetes/pki/*.crt; do
     printf '%s: %s\n' \
     "$(date --date="$(openssl x509 -enddate -noout -in "$crt"|cut -d= -f 2)" --iso-8601)" \
     "$crt"
done | sort

You can also find an opensource workflow for Action Orchestrator that will monitor this daily and alert them to issues.

https://github.com/cisco-cx-workflows/cx-ao-shared-workflows/tree/master/CCSCheckKubernetesExpiration__definition_workflow_01E01VIRWZDE24mWlsHrqCGB9xUix0f9ZxG

Solution

New certificates have to be re-issued via Kubeadm across the cluster and then you need to re-join the worker nodes to the masters.

1. Log in to a master node.
   
   
2. Get its IP address via ip addres show.
   
   

   [root@cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a3 kubernetes]# ip address show
   1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
   link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
   inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
   2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 8920 qdisc pfifo_fast state UP group default qlen 1000
   link/ether fa:16:3e:19:63:a2 brd ff:ff:ff:ff:ff:ff
   inet 192.168.1.20/24 brd 192.168.1.255 scope global dynamic eth0
   valid_lft 37806sec preferred_lft 37806sec
   3: docker0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN group default
   link/ether 02:42:d0:29:ce:5e brd ff:ff:ff:ff:ff:ff
   inet 172.17.0.1/16 scope global docker0
   valid_lft forever preferred_lft forever
   13: tunl0@NONE: <NOARP,UP,LOWER_UP> mtu 1430 qdisc noqueue state UNKNOWN group default qlen 1000
   link/ipip 0.0.0.0 brd 0.0.0.0
   inet 172.16.176.128/32 brd 172.16.176.128 scope global tunl0
   valid_lft forever preferred_lft forever
   14: cali65453a0219d@if4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1430 qdisc noqueue state UP group default
   link/ether ee:ee:ee:ee:ee:ee brd ff:ff:ff:ff:ff:ff link-netnsid 4

3. Navigate to the Kubernetes directory via cd /etc/kubernetes.
   
   
4. Create a file called kubeadmCERT.yaml via vi kubeadmCERT.yaml.
   
   
5. The file should look like this:
   
   

   apiVersion: kubeadm.k8s.io/v1alpha1
   kind: MasterConfiguration
   api:
     advertiseAddress: <IP ADDRESS FROM STEP 2>
   kubernetesVersion: v1.11.6
   #NOTE: If the customer is running a load balancer VM then you must add these lines after...
   #apiServerCertSANs:
   #- <load balancer IP>

6. Backup your old certificates and keys. This is not required but recommended. Make a backup directory and copy these files to it.
   
   

   #Files
   #apiserver.crt
   #apiserver.key
   #apiserver-kubelet-client.crt
   #apiserver-kubelet-client.key
   #front-proxy-client.crt
   #front-proxy-client.key
   
   #ie
   cd /etc/kubernetes/pki
   mkdir backup
   mv apiserver.key backup/apiserver.key.bak

7. If you skipped Step 6., you can just delete the previously mentioned files via rm command like rm apiserver.crt.
   
   
8. Navigate back to where your kubeadmCERT.yaml file is located. Generate a new apiserver cert via kubeadm --config kubeadmCERT.yaml alpha phase certs apiserver.
   
   

   [root@cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a3 kubernetes]# kubeadm --config kubeadmCERT.yaml alpha phase certs apiserver
   [certificates] Generated apiserver certificate and key.
   [certificates] apiserver serving cert is signed for DNS names [cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a3 kubernetes kubernetes.default kubernetes.default.svc kubernetes.default.svc.cluster.local] and IPs [10.96.0.1 192.168.1.20]

9. Generate new apiserver kubelet cert via kubeadm --config kubeadmCERT.yaml alpha phase certs apiserver-kubelet-client.
   
   
10. Generate new front-proxy-client cert via kubeadm --config kubeadmCERT.yaml alpha phase certs front-proxy-client.
    
    
11. In the /etc/kubernetes folder, backup the .conf files. Not required but recommended. You should have kubelet.conf, controller-manager.conf, scheduler.conf, and possibly admin.conf. You can delete them if you don't want to back them up.
    
    
12. Generate new configuration files via kubeadm --config kubeadmCERT.yaml alpha phase kubeconfig all.
    
    

    [root@cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a3 kubernetes]# kubeadm --config kubeadmCERT.yaml alpha phase kubeconfig all
    [kubeconfig] Wrote KubeConfig file to disk: "/etc/kubernetes/admin.conf"
    [kubeconfig] Wrote KubeConfig file to disk: "/etc/kubernetes/kubelet.conf"
    [kubeconfig] Wrote KubeConfig file to disk: "/etc/kubernetes/controller-manager.conf"
    [kubeconfig] Wrote KubeConfig file to disk: "/etc/kubernetes/scheduler.conf"

13. Export your new admin.conf file to your host.
    
    

    cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
    chown $(id -u):$(id -g) $HOME/.kube/config
    chmod 777 $HOME/.kube/config
    export KUBECONFIG=.kube/config

14. Reboot the master node via shutdown -r now.
    
    
15. Once the master is backed up, check whether kubelet is running via systemctl status kubelet.
    
    
16. Verify Kubernetes via kubectl get nodes.
    
    

    [root@cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a3 ~]# kubectl get nodes
    NAME                                                       STATUS     ROLES     AGE       VERSION
    cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a1   Ready      master    1y        v1.11.6
    cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a2   Ready      master    1y        v1.11.6
    cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a3   Ready      master    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a1   NotReady   <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a2   NotReady   <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a3   NotReady   <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a4   NotReady   <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a5   NotReady   <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a6   NotReady   <none>    1y        v1.11.6
    

    
    
17. Repeat steps 1. to 16. for each master node.
    
    
18. On one master, generate a new join token via kubeadm token create --print-join-command. Copy that command for later use.
    
    

    [root@cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a1 k8s-mgmt]# kubeadm token create 
     --print-join-command
    kubeadm join 192.168.1.14:6443 --token m1ynvj.f4n3et3poki88ry4
     --discovery-token-ca-cert-hash 
    sha256:4d0c569985c1d460ef74dc01c85740285e4af2c2369ff833eed1ba86e1167575

19. Get the IPs of your workers via kubectl get nodes -o wide.
    
    
20. Log in to a worker like ssh -i /home/cloud-user/keys/gen3-ao-prod.key cloud-user@192.168.1.17 and navigate to root access.
    
    
21. Stop the kubelet service via systemctl stop kubelet.
    
    
22. Remove the old configuration files, these include ca.crt, kubelet.conf, and bootstrap-kubelet.conf.
    
    

    rm /etc/kubernetes/pki/ca.crt
    rm /etc/kubernetes/kubelet.conf
    rm /etc/kubernetes/bootstrap-kubelet.conf

23. Grab the name of the node from Step 19.
    
    
24. Issue the command for the worker in order to rejoin the cluster. Use the command from 18., but add --node-name <name of node> to the end.
    
    

    [root@cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a1 kubernetes]# kubeadm join 192.168.1.14:6443 --token m1ynvj.f4n3et3poki88ry4 --discovery-token-ca-cert-hash sha256:4d0c569985c1d460ef74dc01c85740285e4af2c2369ff833eed1ba86e1167575 --node-name cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a1
    [preflight] running pre-flight checks
            [WARNING RequiredIPVSKernelModulesAvailable]: the IPVS proxier will not be used, because the following required kernel modules are not loaded: [ip_vs_rr ip_vs_wrr ip_vs_sh] or no builtin kernel ipvs support: map[ip_vs:{} ip_vs_rr:{} ip_vs_wrr:{} ip_vs_sh:{} nf_conntrack_ipv4:{}]
    you can solve this problem with following methods:
     1. Run 'modprobe -- ' to load missing kernel modules;
    2. Provide the missing builtin kernel ipvs support
    
    I0226 17:59:52.644282   19170 kernel_validator.go:81] Validating kernel version
    I0226 17:59:52.644421   19170 kernel_validator.go:96] Validating kernel config
    [discovery] Trying to connect to API Server "192.168.1.14:6443"
    [discovery] Created cluster-info discovery client, requesting info from "https://192.168.1.14:6443"
    [discovery] Requesting info from "https://192.168.1.14:6443" again to validate TLS against the pinned public key
    [discovery] Cluster info signature and contents are valid and TLS certificate validates against pinned roots, will use API Server "192.168.1.14:6443"
    [discovery] Successfully established connection with API Server "192.168.1.14:6443"
    [kubelet] Downloading configuration for the kubelet from the "kubelet-config-1.11" ConfigMap in the kube-system namespace
    [kubelet] Writing kubelet configuration to file "/var/lib/kubelet/config.yaml"
    [kubelet] Writing kubelet environment file with flags to file "/var/lib/kubelet/kubeadm-flags.env"
    [preflight] Activating the kubelet service
    [tlsbootstrap] Waiting for the kubelet to perform the TLS Bootstrap...
    [patchnode] Uploading the CRI Socket information "/var/run/dockershim.sock" to the Node API object "cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a1" as an annotation
    
    This node has joined the cluster:
    * Certificate signing request was sent to master and a response
      was received.
    * The Kubelet was informed of the new secure connection details.
    
    Run 'kubectl get nodes' on the master to see this node join the cluster.
    

25. Exit the worker and check the status on a master via kubectl get nodes. It should be in Ready status.
    
    
26. Repeat steps 20. to 25. for each worker.
    
    
27. The last kubectl get nodes should show that all nodes are in "Ready" status, back online, and joined to the cluster.
    
    

    [root@cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a1 ~]# kubectl get nodes
    NAME                                                       STATUS    ROLES     AGE       VERSION
    cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a1   Ready     master    1y        v1.11.6
    cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a2   Ready     master    1y        v1.11.6
    cx-ccs-prod-master-d7f34f25-f524-4f90-9037-7286202ed13a3   Ready     master    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a1   Ready     <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a2   Ready     <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a3   Ready     <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a4   Ready     <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a5   Ready     <none>    1y        v1.11.6
    cx-ccs-prod-worker-d7f34f25-f524-4f90-9037-7286202ed13a6   Ready     <none>    1y        v1.11.6