tcpdump captures unexpected packets on secondary(backup) bonding slave
Issue
- When taking tcpdump on each bonding slave, tcpdump captures ping ICMP request/receive packets on both primary and secondary bonding slaves which are in active-backup mode.
- The
seq=1
and same id packets can be detected on both slaves and then the subsequent packets are only detected on the primary bonding slave. - Why are the first packets detected on secondary while primary is working?
bond0
14:34:15.208148 IP 10.10.10.10 > 10.10.10.20: ICMP echo request, id 46585, seq 1, length 64
14:34:15.208584 IP 10.10.10.20 > 10.10.10.10: ICMP echo reply, id 46585, seq 1, length 64
14:34:16.211336 IP 10.10.10.10 > 10.10.10.20: ICMP echo request, id 46585, seq 2, length 64
14:34:16.211717 IP 10.10.10.20 > 10.10.10.10: ICMP echo reply, id 46585, seq 2, length 64
14:34:17.235342 IP 10.10.10.10 > 10.10.10.20: ICMP echo request, id 46585, seq 3, length 64
14:34:17.235686 IP 10.10.10.20 > 10.10.10.10: ICMP echo reply, id 46585, seq 3, length 64
...
slave(secondary)
14:34:15.208186 IP 10.10.10.10 > 10.10.10.20: ICMP echo request, id 46585, seq 1, length 64 <<<---
14:34:15.208586 IP 10.10.10.20 > 10.10.10.10: ICMP echo reply, id 46585, seq 1, length 64 <<<---
slave(active)
14:34:15.208151 IP 10.10.10.10 > 10.10.10.20: ICMP echo request, id 46585, seq 1, length 64 <<<---
14:34:15.208584 IP 10.10.10.20 > 10.10.10.10: ICMP echo reply, id 46585, seq 1, length 64 <<<---
14:34:16.211340 IP 10.10.10.10 > 10.10.10.20: ICMP echo request, id 46585, seq 2, length 64
14:34:16.211717 IP 10.10.10.20 > 10.10.10.10: ICMP echo reply, id 46585, seq 2, length 64
14:34:17.235347 IP 10.10.10.10 > 10.10.10.20: ICMP echo request, id 46585, seq 3, length 64
14:34:17.235686 IP 10.10.10.20 > 10.10.10.10: ICMP echo reply, id 46585, seq 3, length 64
...
Environment
- Red Hat Enterprise Linux 8
- Bonding interface with active-backup mode
- Cisco L2 Switch
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.