Kubernetes Common Errors
1.
ImageBackPullOff
We face this issue when the image is not present in registry or the given image tag is wrong.
Make sure you provide correct registry url, image name and image tag.
We might face authentication failures, when image is being stored in a private registry, make sure to create secret with private registry credentials and add created secret in Kubernetes Deployment File to pull docker image.
2.
CrashLoopBackOff
We face this issue when the process deployed inside container not running then the POD will be moved to CrashLoopBackOff.
POD might be running out of CPU or memory, POD should get enough resources allocated that’s cpu and memory for an application to be up and running, to fix that check in Resources Requests and Resources Limits.
3.
OOM Killed - Out Of Memory
We face this issue when PODs tries to utilise more memory than the limits we have set.
We can resolve it by setting appropriate resource request and resource limit.
4.
POD Status - Pending
When nodes might not be ready and required resources like CPU and Memory may not be available in nodes for the PODs to be up and running.
5.
POD Status - Waiting
POD will be scheduled to a node but POD won’t be running in scheduled node.
We can fix this by providing correct image name, image tag and authentication to registry.
6.
POD will be up and running and application is not accessible.
We can fix this by creating appropriate service.
If service is already created and application is still not accessible, make sure application and service are deployed in same namespace.
7.
POD Status - Evicted
We can resolve this by setting appropriate resource requests and resource limits for the PODs and having enough resources in worker nodes.
