I'm having problems with my Kubernetes cluster. I have a service with a deployment associated. I have several replicas of the pods in this deployment up and running and 2 containers for each deployment.

I'm using a Rolling Updatestrategy for redeploying, with parameters maxSurge : 1 and maxUnavailable: 0. When I change a deployment the pods are changing correctly, adding one new pod and then deleting an old one until the whole deployment is changed.

However, on the switch of the first pod from old to new, I'm having a small outage. This doesn't happend when I'm changing the other pods, that is, when there is at least one new pod up and running.

What can I do to fix this issue? I have a livenessProbe and a readinessProbe but still it looks like the new pod is not ready to receive traffic when it just starts, although the readinessProbe works correctly. I've read that in some applications the pod doesn't close with a SIGTERM as Kubernetes expects and therefore the pod is abruptly killed. I'm wondering if this is the problem but can't be sure. What would be the correct way to handle this in the preStop? Is there something that I could be missing?