Pod abnormal troubleshooting

pod state ：

Pending（ It's up in the air ）	Pod Has been Kubernetes The system accepts , But one or more containers have not been created or run . This phase involves waiting Pod The time to be scheduled and the time to download the image through the network ,
Waiting （ wait for ）	Pod be in Waiting The state container is still running, and it does what it needs to start .
Terminated（ Terminated ）	Pod be in Terminated The state of the container has started to execute and either ends normally or fails for some reason ..
Failed（ Failure ）	Pod All containers in have terminated , And at least one container is terminated for failure . in other words , The container is not 0 The state exits or is terminated by the system .
Unknown（ Unknown ）	For some reason can't get Pod The state of . This is usually due to the fact that Pod Host communication failure .
Eviction（ deportation ）	When an exception occurs in a node , To ensure the availability of the workload ,kubernetes There will be a corresponding mechanism to evict Pod.

Event classes

Back-off restarting failed docker container	Restarting abnormal Docker Containers
fit failure on node: Insufficient cpu	node CPU Insufficient
no nodes available to schedule pods	The cluster resources are insufficient
pod failed to fit in any node	No suitable node is available for the instance
Liveness probe failed	Container health check failed
Error syncing pod, skipping	The container process crashes or exits

pod In the above situation , It can be done by kubectl describe pod -n<namepsaces> <podname> View correspondence event Display information , Solve the problem based on the corresponding error message ;

" Back-off And pod The problem of restarting occurs , Can pass ：

kubectl logs -n<namepsaces> <podname> // View the current pod Standard output log

kubectl logs -n<namepsaces> <podname> -p //// Check out the previous pod Standard output log

Solve the corresponding problems based on the corresponding standard output log .