Networking is broken with latest clearlinux about cloud-native-setup HOT 14 CLOSED

clear@clr-01 ~/clr-k8s-examples $ swupd info
Installed version: 29880
clear@clr-01 ~/clr-k8s-examples $ crio --version
crio version 1.14.1
commit: ""
clear@clr-01 ~/clr-k8s-examples $ /lib/systemd/systemd --version
systemd 241 (241)
+PAM +AUDIT -SELINUX +IMA -APPARMOR -SMACK -SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ -LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=legacy
clear@clr-01 ~/clr-k8s-examples $ kubectl create deployment nginx --image=nginx
deployment.apps/nginx created
clear@clr-01 ~/clr-k8s-examples $ kubectl scale --replicas=3 deployment nginx
deployment.extensions/nginx scaled
clear@clr-01 ~/clr-k8s-examples $ for i in $(kubectl get pod -l app=nginx -o jsonpath='{.items[*].status.podIP}'); do curl -s -o /dev/null -w "%{http_code}\n" $i;  done
200
200
200
clear@clr-01 ~/clr-k8s-examples $ sysctl -ar '\.rp_filter'
net.ipv4.conf.all.rp_filter = 2
net.ipv4.conf.default.rp_filter = 1
net.ipv4.conf.enp0s8.rp_filter = 0
net.ipv4.conf.enp0s9.rp_filter = 0
net.ipv4.conf.flannel/1.rp_filter = 1
net.ipv4.conf.lo.rp_filter = 0

clear@clr-01 ~/clr-k8s-examples $ swupd info
Installed version: 29890
Version URL:       https://cdn.download.clearlinux.org/update
Content URL:       https://cdn.download.clearlinux.org/update
clear@clr-01 ~/clr-k8s-examples $ crio --version
crio version 1.14.2
commit: ""
clear@clr-01 ~/clr-k8s-examples $ /lib/systemd/systemd --version
systemd 242 (242)
+PAM +AUDIT -SELINUX +IMA -APPARMOR -SMACK -SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ -LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=legacy
clear@clr-01 ~/clr-k8s-examples $ kubectl create deployment nginx --image=nginx
deployment.apps/nginx created
clear@clr-01 ~/clr-k8s-examples $ kubectl scale --replicas=3 deployment nginx
deployment.extensions/nginx scaled
clear@clr-01 ~/clr-k8s-examples $ for i in $(kubectl get pod -l app=nginx -o jsonpath='{.items[*].status.podIP}'); do curl -s -o /dev/null -w "%{http_code}\n" $i;  done

^C
clear@clr-01 ~/clr-k8s-examples $ sysctl -ar '\.rp_filter'
net.ipv4.conf.all.rp_filter = 2
net.ipv4.conf.cali2b86df72b07.rp_filter = 1
net.ipv4.conf.caliacd620db126.rp_filter = 1
net.ipv4.conf.default.rp_filter = 1
net.ipv4.conf.enp0s8.rp_filter = 0
net.ipv4.conf.enp0s9.rp_filter = 0
net.ipv4.conf.flannel/1.rp_filter = 1
net.ipv4.conf.lo.rp_filter = 0

rp_filter - INTEGER
    0 - No source validation.
    1 - Strict mode as defined in RFC3704 Strict Reverse Path
        Each incoming packet is tested against the FIB and if the interface
        is not the best reverse path the packet check will fail.
        By default failed packets are discarded.
    2 - Loose mode as defined in RFC3704 Loose Reverse Path
        Each incoming packet's source address is also tested against the FIB
        and if the source address is not reachable via any interface
        the packet check will fail.

    Current recommended practice in RFC3704 is to enable strict mode
    to prevent IP spoofing from DDos attacks. If using asymmetric routing
    or other complicated routing, then loose mode is recommended.

    The max value from conf/{all,interface}/rp_filter is used
    when doing source validation on the {interface}.

    Default value is 0. Note that some distributions enable it
    in startup scripts.

# For IPv4, we rely on the kernel's reverse path filtering to prevent
# workloads from spoofing their IP addresses.
#
# The RPF check for a particular interface is controlled by several
# sysctls:
#
# - ipv4.conf.all.rp_filter is a global override
# - ipv4.conf.default.rp_filter controls the value that is set on a newly
#   created interface
# - ipv4.conf.<interface>.rp_filter controls a particular interface.
#
# The algorithm for combining the global override and per-interface values
# is to take the *numeric* maximum between the two.  The values are:
# 0=off, 1=strict, 2=loose.  "loose" is not suitable for Calico since it
# would allow workloads to spoof packets from other workloads on the same
# host.  Hence, we need the global override to be <=1 or it would override
# the per-interface setting to "strict" that we require.
#
# We bail out rather than simply setting it because setting 2, "loose",
# is unusual and it is likely to have been set deliberately.