I have a GKE cluster with an autoscale node pool.
After adding some pods, the cluster starts autoscale and creates a new node but the old running pods start to crash randomly:
I don't think it's directly related to autoscaling unless some of your old nodes are being removed. The autoscaling is triggered by adding more pods but most likely, there is something with your application or connectivity to external services (db for example). I would check the what's going on in the pod logs:
$ kubectl logs <pod-id-that-is-crashing>
You can also check for any other event in the pods or deployment (if you are using a deployment)
$ kubectl describe deployment <deployment-name>
$ kubectl describe pod <pod-id> -c <container-name>
Hope it helps!
Related
I have kubernets cluster in gcp with docker container runtime. I am trying to change docker container runtime into containerd. Following steps shows what I did.
New node pool added ( nodes with containerd )
drained old nodes
Once I perform above steps I am getting " Pod is blocking scale down because it has local storage " warning message.
You need to add the once annotation to POD so that cluster autoscaler can remove that POD from POD safe to evict.
cluster-autoscaler.kubernetes.io/safe-to-evict": "true"
above annotation, you have to add in into POD.
You can read more at : https://cloud.google.com/kubernetes-engine/docs/how-to/cluster-autoscaler-visibility#cluster-not-scalingdown
NoScaleDown example: You found a noScaleDown event that contains a
per-node reason for your node. The message ID is
"no.scale.down.node.pod.has.local.storage" and there is a single
parameter: "test-single-pod". After consulting the list of error
messages, you discover this means that the "Pod is blocking scale down
because it requests local storage". You consult the Kubernetes Cluster
Autoscaler FAQ and find out that the solution is to add a
"cluster-autoscaler.kubernetes.io/safe-to-evict": "true" annotation to
the Pod. After applying the annotation, cluster autoscaler scales down
the cluster correctly.
For further clarification, you can use this command to update the pod's annotation:
kubectl annotate pod <podname> -n <namespace> "cluster-autoscaler.kubernetes.io/safe-to-evict=true"
Had the same error when using Gitlab + Autodevops + GoogleCloud.
The issue is the cm_acme pods's that are spun up to answer the letsencrypt challenges.
e.g. we have pods like this
cm-acme-http-solver-d2tak
hanging around in our cluster so the cluster won't downsize until these pods are destroyed.
A simple
kubectl get pods -A | grep cm-acme
will list all the pods that need to be destroyed with
kubectl delete pod -n {namespace} {pod name}
I ran kubectl rollout restart deployment.
It created a new pod which is now stuck in Pending state because there are not enough resources to schedule it.
I can't increase the resources.
How do I delete the new pod?
please check if that pod has a Deployment controller (which should be recreating the pod), use:
kubectl get deployments
Then try to delete the Deployment with
Kubectl delete deployment DEPLOYMENT_NAME
Also, I would suggest to check resources allocation on GKE and its usage on your nodes with next command:
kubectl describe nodes | grep -A10 "Allocated resources"
And if you need more resources, try to activate GKE CA (cluster autoscaler) or in case you already have it enabled, then increase the number of nodes on Max value. You can also try to manually add a new node by manually resizing the Nodepool you are using.
If there is an update in the docker image, rolling update strategy will update all the pods one by one in a daemonset, similarly is it possible to restart the pods gracefully without any changes the daemonset config or can it be triggered explicitly?
Currently, I am doing it manually by
kubectl delete pod <pod-name>
One by one until each pod gets into running state.
You could try and use Node maintenance operations:
Use kubectl drain to gracefully terminate all pods on the node while marking the node as unschedulable (with --ignore-daemonsets, from Konstantin Vustin's comment):
kubectl drain $NODENAME --ignore-daemonsets
This keeps new pods from landing on the node while you are trying to get them off.
Then:
Make the node schedulable again:
kubectl uncordon $NODENAME
To trigger restart of all pods managed by deamonset in namespace [namespace_name]:
kubectl rollout restart de -n [namespace_name]
I have 3 nodes in kubernetes cluster. I create a daemonset and deployed it in all the 3 devices. This daemonset created 3 pods and they were successfully running. But for some reasons, one of the pod failed.
I need to know how can we restart this pod without affecting other pods in the daemon set, also without creating any other daemon set deployment?
Thanks
kubectl delete pod <podname> it will delete this one pod and Deployment/StatefulSet/ReplicaSet/DaemonSet will reschedule a new one in its place
There are other possibilities to acheive what you want:
Just use rollout command
kubectl rollout restart deployment mydeploy
You can set some environment variable which will force your deployment pods to restart:
kubectl set env deployment mydeploy DEPLOY_DATE="$(date)"
You can scale your deployment to zero, and then back to some positive value
kubectl scale deployment mydeploy --replicas=0
kubectl scale deployment mydeploy --replicas=1
Just for others reading this...
A better solution (IMHO) is to implement a liveness probe that will force the pod to restart the container if it fails the probe test.
This is a great feature K8s offers out of the box. This is auto healing.
Also look into the pod lifecycle docs.
kubectl -n <namespace> delete pods --field-selector=status.phase=Failed
I think the above command is quite useful when you want to restart 1 or more failed pods :D
And we don't need to care about name of the failed pod.
I want to scale up/down the number of machines to increase/decrease the number of nodes in my Kubernetes cluster. When I add one machine, I’m able to successfully register it with Kubernetes; therefore, a new node is created as expected. However, it is not clear to me how to smoothly shut down the machine later. A good workflow would be:
Mark the node related to the machine that I am going to shut down as unschedulable;
Start the pod(s) that is running in the node in other node(s);
Gracefully delete the pod(s) that is running in the node;
Delete the node.
If I understood correctly, even kubectl drain (discussion) doesn't do what I expect since it doesn’t start the pods before deleting them (it relies on a replication controller to start the pods afterwards which may cause downtime). Am I missing something?
How should I properly shutdown a machine?
List the nodes and get the <node-name> you want to drain or (remove from cluster)
kubectl get nodes
1) First drain the node
kubectl drain <node-name>
You might have to ignore daemonsets and local-data in the machine
kubectl drain <node-name> --ignore-daemonsets --delete-local-data
2) Edit instance group for nodes (Only if you are using kops)
kops edit ig nodes
Set the MIN and MAX size to whatever it is -1
Just save the file (nothing extra to be done)
You still might see some pods in the drained node that are related to daemonsets like networking plugin, fluentd for logs, kubedns/coredns etc
3) Finally delete the node
kubectl delete node <node-name>
4) Commit the state for KOPS in s3: (Only if you are using kops)
kops update cluster --yes
OR (if you are using kubeadm)
If you are using kubeadm and would like to reset the machine to a state which was there before running kubeadm join then run
kubeadm reset
Find the node with kubectl get nodes. We’ll assume the name of node to be removed is “mynode”, replace that going forward with the actual node name.
Drain it with kubectl drain mynode
Delete it with kubectl delete node mynode
If using kubeadm, run on “mynode” itself kubeadm reset
Rafael. kubectl drain does work as you describe. There is some downtime, just as if the machine crashed.
Can you describe your setup? How many replicas do you have, and are you provisioned such that you can't handle any downtime of a single replica?
If the cluster is created by kops
1.kubectl drain <node-name>
now all the pods will be evicted
ignore daemeondet:
2.kubectl drain <node-name> --ignore-daemonsets --delete-local-data
3.kops edit ig nodes-3 --state=s3://bucketname
set max and min value of instance group to 0
4. kubectl delete node
5. kops update cluster --state=s3://bucketname --yes
Rolling update if required:
6. kops rolling-update cluster --state=s3://bucketname --yes
validate cluster:
7.kops validate cluster --state=s3://bucketname
Now the instance will be terminated.
The below command only works if you have a lot of replicas, disruption budgets, etc. - but helps a lot with improving cluster utilization. In our cluster we have integration tests kicked off throughout the day (pods run for an hour and then spin down) as well as some dev-workload (runs for a few days until a dev spins it down manually). I am running this every night and get from ~100 nodes in the cluster down to ~20 - which adds up to a fair amount of savings:
for node in $(kubectl get nodes -o name| cut -d "/" -f2); do
kubectl drain --ignore-daemonsets --delete-emptydir-data $node;
kubectl delete node $node;
done
Remove worker node from Kubernetes
kubectl get nodes
kubectl drain < node-name > --ignore-daemonsets
kubectl delete node < node-name >
When draining a node we can have the risk that the nodes remain unbalanced and that some processes suffer downtime. The purpose of this method is to maintain the load balance between nodes as much as possible in addition to avoiding downtime.
# Mark the node as unschedulable.
echo Mark the node as unschedulable $NODENAME
kubectl cordon $NODENAME
# Get the list of namespaces running on the node.
NAMESPACES=$(kubectl get pods --all-namespaces -o custom-columns=:metadata.namespace --field-selector spec.nodeName=$NODENAME | sort -u | sed -e "/^ *$/d")
# forcing a rollout on each of its deployments.
# Since the node is unschedulable, Kubernetes allocates
# the pods in other nodes automatically.
for NAMESPACE in $NAMESPACES
do
echo deployment restart for $NAMESPACE
kubectl rollout restart deployment/name -n $NAMESPACE
done
# Wait for deployments rollouts to finish.
for NAMESPACE in $NAMESPACES
do
echo deployment status for $NAMESPACE
kubectl rollout status deployment/name -n $NAMESPACE
done
# Drain node to be removed
kubectl drain $NODENAME
There exists some strange behaviors for me when kubectl drain. Here are my extra steps, otherwise DATA WILL LOST in my case!
Short answer: CHECK THAT no PersistentVolume is mounted to this node. If have some PV, see the following descriptions to remove it.
When executing kubectl drain, I noticed, some Pods are not evicted (they just did not appear in those logs like evicting pod xxx).
In my case, some are pods with soft anti-affinity (so they do not like to go to the remaining nodes), some are pods of StatefulSet of size 1 and wants to keep at least 1 pod.
If I directly delete that node (using the commands mentioned in other answers), data will get lost because those pods have some PersistentVolumes, and deleting a Node will also delete PersistentVolumes (if using some cloud providers).
Thus, please manually delete those pods one by one. After deleted, kuberentes will re-schedule the pods to other nodes (because this node is SchedulingDisabled).
After deleting all pods (excluding DaemonSets), please CHECK THAT no PersistentVolume is mounted to this node.
Then you can safely delete the node itself :)