Is the mode 6 traffic will receive load balance relatively according the the slaves interfaces?
OS: ubuntu 18.04.6 LTS; bonding driver: v3.7.1; kernel: 4.15.0-213-generic
I had setted the parameter arp_interval and arp_ip_target, or bond-arp_interval and bond-arp_ip_target, or bond_arp_interval and bond_arp_ip_target, and tested the traffic by iperf3, all traffic from different hosts with different arp records(same dest ip with different HWaddress) always receive by the same slave interface in dest host. Maybe the ARP record does not update frequently in the subnet, so it can’t receive traffic load balancing.
How the bonding driver initiates an ARP reply to the peer for updating the ARP record? I can’t find any working parameters to setting it.
Linux Ethernet Bonding Driver HOWTO :
Receive load balancing is handled by Address Resolution Protocol (ARP) negotiation and table mapping to the relevant group interface.
Hence, peers learn the hardware address
of the bond and the balancing of receive traffic
collapses to the current slave. This is handled by
sending updates (ARP Replies) to all the peers with
their individually assigned hardware address such that
the traffic is redistributed.