Pod Lifecycle - A POD deletion after a timeout, not immediately. Why?

General Discussions
1

Hi all,

My first post here and haven’t found an answer here yet, so I’m free and open to any bashings. :wink:

Since the version 1.27.7 (but actually since some latest OS version coming with AKS), I’ve observed an issue, which I’ve described here: [BUG] Pods Are Not Terminated If An AKS Node Is Stopped · Issue #4093 · Azure/AKS · GitHub

As per Pod Lifecycle | Kubernetes: “If a Node dies, the Pods scheduled to that node are scheduled for deletion after a timeout period”

How long is this timeout? From the testing coming from the AKS issue 4093 it seems it’s around 5 minutes. Which means, that if a node dies, then a pod enters a Terminating (Unknown in fact) state after ~5 minutes, not immediately. If that’s the case, then why is this timeout so long?

Many thanks for any info in advance!

Best regards

Stefan