Failed to schedule pod on AKS

Sai Theja 21

We are facing this issue regularly,

As per Autoscale, Pods has to create and run on available nodes but

Pods are changed to pending state and taking longer time to run, meanwhile it is getting terminated.

May i know what might be the cause,

Note: Nodes are sufficiently available and kept in auto scale.

srbhatta-MSFT 8,576 Reputation points Microsoft Employee

2023-01-23T05:02:51.5433333+00:00
Hello @Sai Theja ,

Welcome to Microsoft QnA.

We would require more details to investigate this issue. It's difficult to say for certain without more information on the potential cause, but below are some common causes to investigate.

Limited resources: If the pods require more resources (such as memory or CPU) than are available on the nodes, they will remain in a pending state until resources become available.

Networking issues: If there are networking issues within the cluster, pods may not be able to communicate with each other or with the Kubernetes API server, which can prevent them from being scheduled.

Image pull failures: If the image for the pod cannot be pulled, the pod will remain in a pending state.

Configuration: In case, the pod is using a specific configuration or a specific resource that is not present on the nodes.

Limited IP address: In case, the IP address for the pods is limited then also it can cause pending state.

It would be helpful to check the events and logs of the failed pods and the nodes on which it should have been running.

Let me know if this helps. Thank you.

1 answer

Sai Theja 21 Reputation points

2023-01-23T12:53:14.26+00:00

Hi @srbhatta-MSFT

Thanks for the reply,

I have verified the logs for the problematic time stamp, it was due to node insufficiency, where as nodes are kept in autoscale mode,

Where can i check the logs for the nodes autoscaler, what is the time period it took to allocate new node on to cluster..?

In logs diagnostics, we have enabled

The following logs are enabled.

Please let me know, where we can find out the Cluster autoscaler logs.
Please sign in to rate this answer.

0 comments No comments
Sign in to comment

Use comments to ask for clarification, additional information, or improvements to the question.

Your answer