Dear all,
Ever since August, we have started receiving the following error https://registry-1.docker.io/v2/: net/http: TLS handshake timeout
during some docker pull
operations.
The error is not consistent and appears at random times - it may fail for several days in a row and then run without issues. However, it seems that when it appears it’s usually about the same Docker images.
Finally, we always use a valid, paid account to login to dockerhub before any pull or push actions.
The error is primarily met on a Kubernetes cluster of ours, where several retries of an image pull need to take place in order for the pods to start. It is also met on some docker-in-docker containers that run within pods, in the same Kubernetes cluster.
We have already disabled all security inspection rules for this flow on the network firewall, but the issue persists.
Docker Version: 24.0.0
Cluster’s container runtime: containerd://1.7.11-k3s2