Skip to content

./iptables-tracer -f "-s 10.0.1.250 -p tcp" can't work at CentOS Linux release 8.5.2111 #229

@BurlyLuo

Description

@BurlyLuo

1.issue details:
when exec ./iptables-tracer -f "-s 10.0.1.250 -p tcp" at the node, the ssh conn will be lost, and the k8s platform also come into abnormal status. esp about the kube-apiserver pods. it continues reboot.

  1. env details:
[root@bpf1 iptables-tracer]$ cat /etc/redhat-release 
CentOS Linux release 8.5.2111
[root@bpf1 iptables-tracer]$ uname -r 
4.18.0-348.7.1.el8_5.x86_64
[root@bpf1 iptables-tracer]$ 

k8s version:
[root@bpf1 iptables-tracer]$ k get nodes -owide 
NAME   STATUS   ROLES           AGE     VERSION   INTERNAL-IP    EXTERNAL-IP   OS-IMAGE         KERNEL-VERSION                CONTAINER-RUNTIME
bpf1   Ready    control-plane   5h54m   v1.27.3   192.168.2.91   <none>        CentOS Linux 8   4.18.0-348.7.1.el8_5.x86_64   containerd://1.6.26
bpf2   Ready    <none>          5h30m   v1.27.3   192.168.2.92   <none>        CentOS Linux 8   4.18.0-348.7.1.el8_5.x86_64   containerd://1.6.26
[root@bpf1 iptables-tracer]$ 

it seems like the iptables issue after the tracer, but only add the LOG trace, not sure why it impact the orig iptables. 

Add: it can work fine at CentOS7 with k8s 1.27.3.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions