cilium: 1.9.4 arping from wrong interface timeout
It could be a successor of https://github.com/cilium/cilium/issues/14693 So we have ens18 - external interface and ens19 - internal one and cilium tries to send arping from external interface into the private network:
2: ens18: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether a6:be:e6:a3:3a:a1 brd ff:ff:ff:ff:ff:ff
inet x.x.173.57/26 brd x.x.173.63 scope global noprefixroute ens18
valid_lft forever preferred_lft forever
inet6 xxx/64 scope link
valid_lft forever preferred_lft forever
3: ens19: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 76:ca:d0:7a:34:a8 brd ff:ff:ff:ff:ff:ff
inet 192.168.202.14/24 brd 192.168.202.255 scope global noprefixroute ens19
valid_lft forever preferred_lft forever
inet6 fe80::74ca:d0ff:fe7a:34a8/64 scope link
valid_lft forever preferred_lft forever
level=error msg="arping failed" error="interrupted system call" interface=ens18 ipAddr=192.168.202.56 subsys=node-neigh
level=error msg="arping failed" error=timeout interface=ens18 ipAddr=192.168.202.53 subsys=node-neigh
root@kube-4.prod.xxx.xxx:~# arping -I ens18 192.168.202.56
ARPING 192.168.202.56 from x.x.173.57 ens18
^CSent 2 probes (2 broadcast(s))
Received 0 response(s)
root@kube-4.prod.xxx.xxx:~# arping -I ens19 192.168.202.56
ARPING 192.168.202.56 from 192.168.202.14 ens19
Unicast reply from 192.168.202.56 [AC:16:2D:BE:3A:AB] 0.891ms
Unicast reply from 192.168.202.56 [AC:16:2D:BE:3A:AB] 0.900ms
^CSent 2 probes (1 broadcast(s))
About this issue
- Original URL
- State: closed
- Created 3 years ago
- Comments: 15 (15 by maintainers)
@brb Thx for your support and attention. We’ve decided to set INTERNAL IPs of k8s heads to IPs of the internal network. And everything is OK now and no errors since this moment. Seems like previously we didn’t face this problem due to arping was introduced in Cilium recently (~ 1.9.2). Now all seems OK. Thank you once again.