CoreDNS pods are not rechable for resolution from other node about rke

[maslo64]

After applying fix from projectcalico/calico#3145 DNS resolution works cross nodes.

#ethtool -k  flannel.1 | grep tx-checksum-ip-generic
        tx-checksum-ip-generic: on
#ethtool -K flannel.1 tx-checksum-ip-generic off
Actual changes:
tx-checksum-ip-generic: off
tx-tcp-segmentation: off [not requested]
tx-tcp-ecn-segmentation: off [not requested]
tx-tcp-mangleid-segmentation: off [not requested]
tx-tcp6-segmentation: off [not requested]
#

[kagipra]

Is it reboot proof? I upgraded k8s to latest 1.26.4, and it still reoccurring.

[superseb]

@rancher/rke1-team According to projectcalico/calico#3145 (comment), there is a patched kernel and an environment variable that can be configured to work-around this. Do we need to add this to the next KDM template?

[manuelbuil]

What CNI plugin are you using?

[maslo64]

@manuelbuil We used default installation so canal plugin.

#kubectl get ds -n kube-system
NAME    DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR            AGE
canal   2         2         2       2            2           kubernetes.io/os=linux   14d
#kubectl get pods -A -l k8s-app=canal
NAMESPACE     NAME          READY   STATUS    RESTARTS         AGE
kube-system   canal-rbksn   2/2     Running   6 (3d23h ago)    14d
kube-system   canal-tgkt9   2/2     Running   10 (3d23h ago)   14d
#

I've noticed very similar issue reported for RKE2 https://docs.rke2.io/known_issues/#calico-with-vxlan-encapsulation

@kagipra disabling offloading via ethtool is not reboot persistent.

[iwalmsley]

This is occurring for me as well for a brand new Ubuntu 22 cluster - and the workaround suggested does work.

[github-actions]

This repository uses an automated workflow to automatically label issues which have not had any activity (commit/comment/label) for 60 days. This helps us manage the community issues better. If the issue is still relevant, please add a comment to the issue so the workflow can remove the label and we know it is still valid. If it is no longer relevant (or possibly fixed in the latest release), the workflow will automatically close the issue in 14 days. Thank you for your contributions.