I am currently working on a monitoring service that will monitor Kubernetes' deployments and their pods. I want to notify users when a deployment is not running the expected amount of replicas and also when pods' containers restart unexpectedly. This may not be the right things to monitor and I would greatly appreciate some feedback on what I should be monitoring.
Anyways, the main question is the differences between all of the Statuses of pods. And when I say Statuses I mean the Status column when running kubectl get pods. The statuses in question are:
- ContainerCreating
- ImagePullBackOff
- Pending
- CrashLoopBackOff
- Error
- Running
What causes pod/containers to go into these states?
For the first four Statuses, are these states recoverable without user interaction?
What is the threshold for a CrashLoopBackOff?
Is Running the only status that has a Ready Condition of True?
Any feedback would be greatly appreciated!
Also, would it be bad practice to use kubectl in an automated script for monitoring purposes? For example, every minute log the results of kubectl get pods to Elasticsearch?
You can see the pod lifecycle details in k8s documentation.
The recommended way of monitoring kubernetes cluster and applications are with prometheus
I will try to tell what I see hidden behind these terms
ContainerCreating
Showing when we wait to image be downloaded and the
container will be created by a docker or another system.
ImagePullBackOff
Showing when we have problem to download the image from a registry. Wrong credentials to log in to the docker hub for example.
Pending
The container starts (if start take time) or started but redinessProbe failed.
CrashLoopBackOff
This status showing when container restarts occur too much often. For example, we have process that tries to read not exists file and crash. Then the container will be recreated by Kube and repeat.
Error
This is pretty clear. We have some errors to run the container.
Running
All is good container running and livenessProbe is OK.
Related
I want to set up alerts when any pod in my Kubernetes cluster is in a CrashloopBackOff state. I'm running Kubelet on Azure Kubernetes Services and have set up a Prometheus Operator which exposes metrics/cadvisor.
Other similar questions on this topic, such as this and this are not relevant to Kubelet setups. The recommended kube_pod_container_status_waiting_reason{}/kube_pod_status_phase{phase="Pending|Unknown|Failed"} and similar queries are not available to me with Kubelet on AKS.
Kubelet has somewhat limited metrics, here is what I have tried:
Container state:
container_tasks_state{container='my_container', kubernetes_azure_com_cluster='my_cluster'}
This seems like it should be the right solution, but the state is always 0, whether Running or in CrashloopBackOff. This seems to be a known bug.
Time from start:
time() - container_start_time_seconds{kubernetes_azure_com_cluster='my_cluster', container='my_container'}
We can here notify when the time the container is live is low. Any pod with a repeat alert is crashing. Inelegant as healthy containers will also notify until they've lived long enough, also my alert channel becomes very noisy.
Detect exited containers:
kubelet_running_containers{kubernetes_azure_com_cluster='my_cluster', container_state='exited'}
Can detect a crashing container, but containers may also exit gracefully, so a notification on container exits is not very useful. We essentially get a 'container exited' alert and then need to manually check whether it was a crash or graceful exit.
Number of running pods:
kubelet_running_pods{kubernetes_azure_com_cluster='my_cluster'}
Does not change on a container crash.
Scrape error:
container_scrape_error{kubernetes_azure_com_cluster='my_cluster'}
Again, does not change on a container crash.
Which query will allow me to discover if a pod has entered the CrashloopBackOff state?
I have a job I am running in Iguazio. It starts and then the status is "Pending" and the icon is blue. It stays like this indefinitely and there is nothing in the logs that describes what is going on. How do I fix this?
A job stuck in this status is usually a Kubernetes issue. The reason there is no logs in the Iguazio dashboard for the job is because the pod never started, which is where the logs come from. You can navigate to the web shell / Jupyter service in Iguazio and use kubectl commands to find out what is going on in Kubernetes. Usually, I see this when there is an issue with the docker image for the pod, it either can’t be found or has bugs.
In a terminal: doing kubectl get pods and find your pod. It usually has ImagePullBackOff, or CrashLoopBackOff or some similar error. Check the docker image which is usually the culprit. You can kill the pod in Kubernetes, which in turn will error the job out. You can also “abort” the job from the menu in the dashboard under that specific job.
Is there a way to see why a kubernetes pod is failing with the status "craskLoopBackOff" under a heavy load?
I have a HorizontalPodAutoscaler which never kicks in. In its status it always shows low (Under 50%) cpu and memory usage.
Tailing the application logs within the pods doesnt give any insights either.
Try looking at the Kubernetes events kubectl get events --sort-by='.lastTimestamp'
If you don't get anything meaningful out of events go to the specific node and see the kubelet logs journalctl -u kubelet
To get logs from a pod you should use:
kubectl logs [podname] -p
You can also do kubelet logs but that's mostly for Cluster logs.
If there is no logs that means your application did not produces any logs before the crash. You would need to rewrite the app and for example add a memory dump on crush.
You mentioned that the pod is dying under heavy load but stats shows only 50% utilization. You should login to the pod and check yourself the load, maybe check how many files are being open because maybe you are hitting the limit.
You can read the Kubernetes docs about Application Introspection and Debugging and go over Debugging CrashLoopBackoffs with Init-Containers.
You can also try running your image in Docker and checking logs there. There is a nice documentation about Logs and troubleshooting available.
If you provide more details we might be more helpful.
Below are some obvious reasons for crashloopbackoff, which I have observed:
waiting for some condition to be full-filled e.g. some secrets,
failing healthcheck etc
pod is running with burstable or besteffort
QoS and is getting killed due to non-availability of resources on
node
You can run this script to find the possible issues for pods in a namespace: https://github.com/dguyhasnoname/k8s-day2-ops/blob/master/namespace_scripts/debug_app_namespace.sh
I am new to Kubernetes and started working with it from past one month.
When creating the setup of cluster, sometimes I see that Heapster will be stuck in Container Creating or Pending status. After this happens the only way have found here is to re-install everything from the scratch which has solved our problem. Later if I run the Heapster it would run without any problem. But I think this is not the optimal solution every time. So please help out in solving the same issue when it occurs again.
Heapster image is pulled from the github for our use. Right now the cluster is running fine, So could not send the screenshot of the heapster failing with it's status by staying in Container creating or Pending status.
Suggest any alternative for the problem to be solved if it occurs again.
Thanks in advance for your time.
A pod stuck in pending state can mean more than one thing. Next time it happens you should do 'kubectl get pods' and then 'kubectl describe pod '. However, since it works sometimes the most likely cause is that the cluster doesn't have enough resources on any of its nodes to schedule the pod. If the cluster is low on remaining resources you should get an indication of this by 'kubectl top nodes' and by 'kubectl describe nodes'. (Or with gke, if you are on google cloud, you often get a low resource warning in the web UI console.)
(Or if in Azure then be wary of https://github.com/Azure/ACS/issues/29 )
I have a kubernetes cluster running fine. It has 4 workers and 1 master with the dashboard to view the status. After running it for sometime, I looked at the Restart count of a node and it was 8. I immediately ran the describe command to get any events but there was no events for that pod. However when I checked the logs of the containers, I found out that the node itself was powered down and up 4 times but dont know why it didnt had any events.
In another node, while looking at the restart count, I got event as Sandbox changed which means probably the node was powered down for sometime and thus the master lost connection to it and so incremented the restart count by 2.
I wanted to know how can we get the logs/debug related to this restart count to know why it was restarted.
Whenever a pod is recreated, does it takes up a new name.? If so, how can we get the events of the previous pod.
Does sandbox changed event actually means that master actually lost connection.?
Step by step:
I'd check the kubelet and docker daemon logs, these restarts should appear somewhere in the logs and hopefully more info about what causes them.
Yes, the pod's name is unique thus it change everytime a pod is destroyed and recreated. You can try to find the pod with kubectl get po -a. Other solution is to get all events with kubectl get events and then filter to find your pod's events.
I've seen this error before and in my case it meant problem with the docker daemon networking. But I searched a bit in google and I saw many other reasons. Again, try to analyse the docker daemon and kubelet logs, and also dmesg. If you have doubts please add a link to the logs in your question and I'll try to help.