Shulou(Shulou.com)05/31 Report--

This article shows you what are the five simple tips for solving Kubernetes Pod problems, which are concise and easy to understand, which will definitely brighten your eyes. I hope you can get something through the detailed introduction of this article.

In many cases, you may find that applications in Kubernetes are not deployed correctly or are not working properly. Today we will provide some tips and tips on how to solve this kind of problem quickly. In addition, I will share with you some very useful tips on how to operate Kubernetes.

So, let's get started!

First of all, there are generally two reasons why Pod fails:

Errors in Kubernetes resource configuration, such as in deployment (Deployment) and service (Service).

Problems in the code.

In the first case, the container usually does not start. In the latter example, the application code fails after the container starts. We will deal with each situation systematically.

In this exercise, we will use kubectl to interact with Kubernetes.

Tip 1: check Pod

Confirm that the Pod is in the Running state or Ready state.

Kubectl get pods

As shown in the picture, a Pod is waiting (Pending) for 9 hours, which is definitely not a good thing! The container doesn't start, and we'll take a closer look at it using the describe command in Tip 2. Here, however, we highlight other error codes that occur when the container fails to start. As follows:

Imagepullbackoff:

The Docker image repository is not accessible and the image name or version specified in the deployment is incorrect.

Make sure that the image name is correct and that the image repository is accessible and authenticated (docker login … ).

RunContainerError:

It's also a possibility.

Reason:

ConfigMap or Secrets is missing.

ContainerCreating:

Some components cannot be enabled immediately when the container is created, such as persistent volumes.

Before examining the other errors, let's try to start Pod with the wrong image name.

# start Pod from image "ngin". # 'web' can be any name, is the name of resulting K8S deployment kubectl run web-image=ngin-replicas=1

The last line shows the mirror error

Sure enough, the completely non-existent mirror "ngin" caused the ImagePullBackOff error. Using the correct mirror name "nginx" can solve this problem.

Kubectl run temp-image=nginx-replicas=1 kubectl get pods

In the picture, Pod is up.

Next, here are some errors that may occur after the container starts.

Crashloopbackoff:

Pod Survivability check failed or Docker mirror error.

For example, Docker CMD exits immediately.

You can use the following tip 3 to check the log.

Note:

The "RESTARTS" column in this screenshot shows the number of restarts.

In this case, you should see some reboots, because when an error occurs, Kubernetes will repeatedly try to start Pod.

If Pod is Running and your application is still not working properly, continue with tips 3 and 4.

Tip 2: check for events related to Pod

If you see an error code on the Pod status, you can use the describe command for more information. This is helpful when the container itself is not started.

Kubectl describe frontend-65c58c957d-f4cqn

The last line of the screenshot shows that Pod has not been started due to the lack of CPU resources, see the message at the bottom. You can add Pod's CPU and redeploy the application.

Tip 3: check the log (Log)

Now that the container is started, you can check the log to see if the application is running properly. For example, Pod frontend-65c58c957d-bzbg2:

Kubectl logs-tail=10 frontend-65c58c957d-bzbg2

Scroll to view a running log in real time:

Kubectl logs-f frontend-65c58c957d-bzbg2

If you don't have any output after kubectl logs, try using get pod, and you'll find that this is most likely a newly started Pod, so you can try to check the logs of some of the last dead Pod.

Kubectl logs frontend-65c58c957d-bzbg2-previous

Tip 4: run "sh", "bash" or "ash" directly in Pod

You can go inside Pod and run commands to troubleshoot the application (type exit to exit).

Kubectl exec-it frontend-65c58c957d-bzbg2 / bin/sh

Tip 5: display cluster-level events

Kubernetes triggers the corresponding event when the state of the resource it manages changes (normal, warning, etc.). This can help us understand what's going on behind it. The get events command provides an aggregate perspective of events.

# all events sorted by time. Kubectl get events-sort-by=.metadata.creationTimestamp# warnings only kubectl get events-field-selector type=Warning# events related to Nodes kubectl get events-field-selector involvedObject.kind=Node

Extra skills

This is my favorite technique! Proficiency in a variety of commands will give you more confidence in walking in the Kubernetes cluster.

First, type kubectl to list all the commands for kubectl.

Next, try to execute the grep debug command with the following command.

Kubectl | grep-I-A 10 debugging

Lists some basic commands that can be run on Kubernetes.

Kubectl | grep-I-A 5 Basic

Next, list the Kubernetes resources that can be manipulated.

Kubectl api-resources

Now you can make some orders by yourself! You can select a command (get, describe, explain) and select a resource and run it! For example, get nodes. So try something else!

While some combinations may not make much sense, apart from this, the entire command system is fairly intuitive and consistent; you can easily write commands and explore.

Just be careful not to delete or modify objects you don't want to encounter.

List the Kubernetes namespaces (namespace):

Kubectl get ns

In this way, you can use specific commands to delve deeper into the corresponding options or examples.

Kubectl get-- help # see K8S system pods in 'kube-system' namespace! Kubectl-n kube-system get pods

As you can see, Kubernetes's command system is very easy to understand, and we can learn a lot by simply testing these commands.

With these, I hope you can find and fix errors in Kubernetes resources and code in the Kubernetes cluster. At the same time, I intend to introduce Kubernetes Services (Service) and network debugging next.

What are the five simple tips for solving Kubernetes Pod problems? have you learned any knowledge or skills? If you want to learn more skills or enrich your knowledge reserve, you are welcome to follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.