Cluster Maintenance

OS Upgrades

When performing upgrades on worker nodes, It's best practice to:

Drain the node kubectl drain node01
When upgrade is completed, you must uncordon it to make it schedulable kubectl uncordon node01

If you only want to make the node unschedulable you can run:

kubectl cordon node01 # Make sure no pods are scheduled on node01

Lab:

Everything went well.

Kubernetes Releases

When you download a new Kubernetes release, it comes with all Kubernetes components with of same version, although some components like etcd and coreDNS come with different versions as they are separate projects.

Releases · kubernetes/kubernetesGitHub

Cluster Upgrade Process

During an upgrade process to the cluster, here are important aspects to note, lets take v1.10 as an example:

If kube-apiserver is at v1.10, then:

controller-manager must be v1.9 or v1.10
kube-scheduler must be v1.9 or v1.10
kubelet must be v1.8, v1.9 or v1.10
kube-proxy must be v1.8, v1.9 or v1.10

This is not the case for kubectl as it can be v1.9, v1.10 or v1.11

These versions allows us to do live upgrade, component by component if required.

When shall we upgrade?

Lets say we are at Kubernetes v1.10, we only have access to the 3 latest version, so if v1.13 is released, v1.10 will be unsupported. So before v1.13 release, it would be a great time to upgrade.

To upgrade, its recommended to upgrade one by one which means v1.10 then v1.11 then v1.12 then v1.13.

Upgrading Process

The Kubernetes upgrade process is dependent on how the cluster is setup.

If we deployed our cluster using AWS EKS, it can be done within a few clicks.
If we deployed our cluster using kubeadm tool, it can be done within commands.
If we deployed our cluster the Kubernetes hard way (From scratch), then we need to manually upgrade components ourselves

We will be upgrading the cluster using kubeadm tool.

If we have cluster with master and worker nodes, and they are at v1.10, there are 2 major steps:

1) Upgrade master nodes: All management functions are down, although worker nodes still serve users

2) Upgrade worker nodes

Upgrading worker nodes has 3 strategies:

1) Bring all nodes down then up (Has downtime)

2) One node at a time

3) Creating new nodes with newer versions and deleting old nodes (Convenient if on cloud)

Kubeadm - Upgrade

We can see components versions and additional useful information if we run:

kubeadm upgrade plan

Remember, kubeadm doesn't install or upgrade kubelets, so we need to deal with it.

Also, we need to upgrade the kubeadm tool itself before we perform the cluster upgrade.

Through this page in the k8s docs, you can upgrade the cluster step by step, starting with the control plane:

Upgrading kubeadm clustersKubernetes

I will share my experience when trying to set it up using the docs in the upcoming lab.

Lab (IMPORTANT):

Now, I will discuss what happened through my lab, I ran kubeadm upgrade plan which appeared to me that my cluster version is v1.28.0 and got the target upgrade that I can upgrade to is v1.28.7 although the question asked to upgrade to exactly v1.29.0.

I had to change the package repository of kubernetes to v1.29 by editing /etc/apt/sources.list.d/kubernetes.list file.

After changing the Kubernetes package repository to v1.29, I ran:

sudo apt update
sudo apt-cache madison kubeadm

These commands gave me the versions I can upgrade my kubdeadm to which gave me v1.29.0-1.1 which I wanted.

I proceeded and used the =1.29.0-1.1 throughout the upgrade commands until I finished upgrading controlplane and uncordoned it.

After that, I had to upgrade worker nodes, the huge mistake I did is trying to run kubectl and kubeadm commands in worker node, it won't work as this is not a master node and doesn't have the components of controlplane.

So I had to:

kubectl drain node01 --ignore-daemonsets # Drain the worker node
ssh node01 # Enter the worker node

After entering the worker node, all I had to do is run the same commands as master node upgrade with the =1.29.0-1.1 but I had to add one command before upgrading kubeadm which is sudo kubeadm upgrade node as shown in k8s worker node upgrade docs

For example:

sudo kubeadm upgrade node
sudo apt-mark unhold kubeadm && \
sudo apt-get update && sudo apt-get install -y kubeadm='1.29.0-1.1' && \
sudo apt-mark hold kubeadm

And then upgrade kubelet and kubectl:

sudo apt-mark unhold kubelet kubectl && \
sudo apt-get update && sudo apt-get install -y kubelet='1.29.0-1.1' kubectl='1.29.0-1.1' && \
sudo apt-mark hold kubelet kubectl

Then restart the kubelet daemon:

sudo systemctl daemon-reload
sudo systemctl restart kubelet

Now, our work is done, so we can exit the worker node and uncordon it from the control plane:

exit
kubectl uncordon node01

That's it!

Backup & Restore Methods

Backup - Resource Configs

One way to backup all services is to query kube-apiserver or use the kubectl utility:

kubectl get all --all-namespaces -o yaml > all-deploy-services.yaml

This is only for few resources groups, there are other resource groups which must be considered

Some solutions such as velero take care of that for you.

Backup - ETCD

All resources created within the cluster is stored in ETCD server, so its crucial to take a backup for it.

The ETCD server is hosted on master nodes, when configuring ETCD we specify a path where all data would be stored.

We could also take a snapshot from the ETCD server, which is a feature built-in in the ETCD server:

ETCDCTL_API=3 etcdctl snapshot save snapshot.db # Or full path (/var/snapshot.db)

ETCDCTL_API=3 etcdctl snapshot status snapshot.db # Status of backup

Remember to specify endpoints, cacert, etcd cert and key for authentication and access to the etcd cluster when using snapshot save.

Restore - ETCD

To restore an ETCD snapshot, we must first stop the kube-apiserver because kube-apiserver depends on ETCD:

service kube-apiserver stop

ETCDCTL_API=3 etcdctl snapshot restore snapshot.db --data-dir /var/lib/etcd-from-backup

Then we need to reload daemon and etcd services:

systemctl daemon-reload
service etcd restart

Finally, start the kube-apiserver:

service kube-apiserver start

If using managed kubernetes cluster, you might not have access to etcd server, so resource backups using the kube-apiserver is better.

Lab 1 (IMPORTANT):

First Restore the snapshot:

ETCDCTL_API=3 etcdctl snapshot restore /opt/snapshot-pre-boot.db --data-dir /var/lib/etcd-from-backup

In this case, we are restoring the snapshot to a different directory but in the same server where we took the backup (the control plane node). As a result, the only required option for the restore command is the --data-dir

Next, update the `/etc/kubernetes/manifests/etcd.yaml`:

We have now restored the etcd snapshot to a new path on the controlplane - /var/lib/etcd-from-backup, so, the only change to be made in the YAML file, is to change the hostPath for the volume called etcd-data from old directory (/var/lib/etcd) to the new directory (/var/lib/etcd-from-backup).

With this change, /var/lib/etcd on the container points to /var/lib/etcd-from-backup on the control plane (which is what we want).

When this file is updated, the ETCD pod is automatically re-created as this is a static pod placed under the /etc/kubernetes/manifests directory.

Note 1: As the ETCD pod has changed it will automatically restart, and also kube-controller-manager, kube-apiserver and kube-scheduler. Wait 1-2 to mins for this pods to restart. You can run the command: watch "crictl ps | grep etcd" to see when the ETCD pod is restarted.

Note 2: If the etcd pod is not getting Ready 1/1, then restart it by kubectl delete pod -n kube-system etcd-controlplane and wait 1 minute.

Note 3: This is the simplest way to make sure that ETCD uses the restored data after the ETCD pod is recreated. You don't have to change anything else.

THIS STEP IS OPTIONAL AND NOT NEEDED FOR COMPLETING THE RESTORE

If you do change --data-dir to /var/lib/etcd-from-backup in the ETCD YAML file, make sure that the volumeMounts for etcd-data is updated as well, with the mountPath pointing to /var/lib/etcd-from-backup

Lab 2 (IMPORTANT):

Most important things to note in this lab:

When switching to a new cluster and you want to check components pods, make sure to ssh to the controlplane of the cluster.
When you can't see an etcd pod, it's most probably an external etcd and you must describe the kube-apiserver to see the external etcd IP, it must be mentioned there so you can ssh into it if needed.
Running commands like ps -aux | grep etcd after ssh into the etcd server is crucial to get information regarding endpoints, data dirs, certs etc...
Make sure to utilize the scp command to copy files from a node to another node specially when doing etcd snapshot restore and backup, examples shown below.
Make sure to also utilize chown specially when restoring etcd snapshot.
Make sure to edit the etcd service to update the data-dir after restoring, usually found at /etc/systemd/system/etcd.service path.

Never edit a static pod directly if you want to modify a data-dir for instance in etcd pod, always check the /etc/kubernetes/manifests folder, if not found then check the /etc/systemd/system/etcd.service (External etcd).

kubelets always re-creates static pods if deleted. So to apply changes delete the static pod!

On the `student-node`:

First set the context to cluster1:

student-node ~ ➜  kubectl config use-context cluster1
Switched to context "cluster1".

Next, inspect the endpoints and certificates used by the etcd pod. We will make use of these to take the backup.

student-node ~ ✖ kubectl describe  pods -n kube-system etcd-cluster1-controlplane  | grep advertise-client-urls
      --advertise-client-urls=https://10.1.218.16:2379

student-node ~ ➜  

student-node ~ ➜  kubectl describe  pods -n kube-system etcd-cluster1-controlplane  | grep pki
      --cert-file=/etc/kubernetes/pki/etcd/server.crt
      --key-file=/etc/kubernetes/pki/etcd/server.key
      --peer-cert-file=/etc/kubernetes/pki/etcd/peer.crt
      --peer-key-file=/etc/kubernetes/pki/etcd/peer.key
      --peer-trusted-ca-file=/etc/kubernetes/pki/etcd/ca.crt
      --trusted-ca-file=/etc/kubernetes/pki/etcd/ca.crt
      /etc/kubernetes/pki/etcd from etcd-certs (rw)
    Path:          /etc/kubernetes/pki/etcd

student-node ~ ➜

SSH to the controlplane node of cluster1 and then take the backup using the endpoints and certificates we identified above:

cluster1-controlplane ~ ➜  ETCDCTL_API=3 etcdctl --endpoints=https://10.1.220.8:2379 --cacert=/etc/kubernetes/pki/etcd/ca.crt --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key snapshot save /opt/cluster1.db
Snapshot saved at /opt/cluster1.db

cluster1-controlplane ~ ➜

Finally, copy the backup to the student-node. To do this, go back to the student-node and use scp as shown below:

student-node ~ ➜  scp cluster1-controlplane:/opt/cluster1.db /opt
cluster1.db                                                                                                        100% 2088KB 112.3MB/s   00:00    

student-node ~ ➜

Step 1

Copy the snapshot file from the student-node to the etcd-server. In the example below, we are copying it to the /root directory:

student-node ~  scp /opt/cluster2.db etcd-server:/root
cluster2.db                                                                                                        100% 1108KB 178.5MB/s   00:00    

student-node ~ ➜

Step 2

Restore the snapshot on the cluster2. Since we are restoring directly on the etcd-server, we can use the endpoint https:/127.0.0.1. Use the same certificates that were identified earlier. Make sure to use the data-dir as /var/lib/etcd-data-new:

etcd-server ~ ➜  ETCDCTL_API=3 etcdctl --endpoints=https://127.0.0.1:2379 --cacert=/etc/etcd/pki/ca.pem --cert=/etc/etcd/pki/etcd.pem --key=/etc/etcd/pki/etcd-key.pem snapshot restore /root/cluster2.db --data-dir /var/lib/etcd-data-new
{"level":"info","ts":1662004927.2399247,"caller":"snapshot/v3_snapshot.go:296","msg":"restoring snapshot","path":"/root/cluster2.db","wal-dir":"/var/lib/etcd-data-new/member/wal","data-dir":"/var/lib/etcd-data-new","snap-dir":"/var/lib/etcd-data-new/member/snap"}
{"level":"info","ts":1662004927.2584803,"caller":"membership/cluster.go:392","msg":"added member","cluster-id":"cdf818194e3a8c32","local-member-id":"0","added-peer-id":"8e9e05c52164694d","added-peer-peer-urls":["http://localhost:2380"]}
{"level":"info","ts":1662004927.264258,"caller":"snapshot/v3_snapshot.go:309","msg":"restored snapshot","path":"/root/cluster2.db","wal-dir":"/var/lib/etcd-data-new/member/wal","data-dir":"/var/lib/etcd-data-new","snap-dir":"/var/lib/etcd-data-new/member/snap"}

etcd-server ~ ➜

Step 3

Update the systemd service unit file for etcdby running vim /etc/systemd/system/etcd.service and add the new value for data-dir:

[Unit]
Description=etcd key-value store
Documentation=https://github.com/etcd-io/etcd
After=network.target

[Service]
User=etcd
Type=notify
ExecStart=/usr/local/bin/etcd \
  --name etcd-server \
  --data-dir=/var/lib/etcd-data-new \
---End of Snippet---

Step 4

Make sure the permissions on the new directory is correct (should be owned by etcd user)

etcd-server /var/lib ➜  chown -R etcd:etcd /var/lib/etcd-data-new

etcd-server /var/lib ➜ 


etcd-server /var/lib ➜  ls -ld /var/lib/etcd-data-new/
drwx------ 3 etcd etcd 4096 Sep  1 02:41 /var/lib/etcd-data-new/
etcd-server /var/lib ➜

Step 5

Finally, reload and restart the etcd service.

etcd-server ~ ➜  systemctl daemon-reload 
etcd-server ~ ➜  systemctl restart etcd
etcd-server ~ ➜

Working with ETCDCTL

etcdctl is a command line client for etcd.

In all our Kubernetes Hands-on labs, the ETCD key-value database is deployed as a static pod on the master. The version used is v3.

To make use of etcdctl for tasks such as back up and restore, make sure that you set the ETCDCTL_API to 3.

You can do this by exporting the variable ETCDCTL_API prior to using the etcdctl client.

This can be done as follows:

export ETCDCTL_API=3

To see all the options for a specific sub-command, make use of the -h or --help flag.

Example:

ETCDCTL_API=3 etcdctl snapshot save -h
ETCDCTL_API=3 etcdctl snapshot restore -h

Slides

2MB

Kubernetes-CKA-0500-Cluster+Maintenance-v1.2.pdf

PDF

Open

PreviousApplication Lifecycle Management NextSecurity

Last updated 1 year ago

OS Upgrades

If you only want to make the node unschedulable you can run:

Lab:

Kubernetes Releases

Cluster Upgrade Process

If kube-apiserver is at v1.10, then:

When shall we upgrade?

Upgrading Process

If we have cluster with master and worker nodes, and they are at v1.10, there are 2 major steps:

Upgrading worker nodes has 3 strategies:

Kubeadm - Upgrade

We can see components versions and additional useful information if we run:

Through this page in the k8s docs, you can upgrade the cluster step by step, starting with the control plane:

Lab (IMPORTANT):

After changing the Kubernetes package repository to v1.29, I ran:

So I had to:

For example:

And then upgrade kubelet and kubectl:

Then restart the kubelet daemon:

Now, our work is done, so we can exit the worker node and uncordon it from the control plane:

Backup & Restore Methods

Backup - Resource Configs

One way to backup all services is to query kube-apiserver or use the kubectl utility:

Backup - ETCD

We could also take a snapshot from the ETCD server, which is a feature built-in in the ETCD server:

Restore - ETCD

To restore an ETCD snapshot, we must first stop the kube-apiserver because kube-apiserver depends on ETCD:

Then we need to reload daemon and etcd services:

Finally, start the kube-apiserver:

Lab 1 (IMPORTANT):

First Restore the snapshot:

Next, update the /etc/kubernetes/manifests/etcd.yaml:

THIS STEP IS OPTIONAL AND NOT NEEDED FOR COMPLETING THE RESTORE

Lab 2 (IMPORTANT):

Most important things to note in this lab:

On the student-node:

Step 1

Step 2

Step 3

Step 4

Step 5

Working with ETCDCTL

This can be done as follows:

Example:

Slides

Next, update the `/etc/kubernetes/manifests/etcd.yaml`:

On the `student-node`: