Bonding of bonded interfaces

RouterOS General
1

Hi all,

I have the following situation:

  • CCR1016-12G running 6.34.2
  • 2 LAGs (bonding1, bonding2) with four 1 GigE links each, connected to two switches.
  • 1 active-backup bonding interface (bonding3) to provide a fault tolerant connection to the switches.

The configuration looks as follows:

Flags: X - disabled, R - running 
 0  R name="bonding1" mtu=1500 mac-address=4C:5E:0C:B1:0B:78 arp=enabled 
      slaves=ether5,ether6,ether7,ether8
      mode=802.3ad primary=none link-monitoring=mii arp-interval=100ms arp-ip-targets="" mii-interval=100ms 
      down-delay=0ms up-delay=0ms lacp-rate=1sec transmit-hash-policy=layer-2 min-links=1 

 1  R name="bonding2" mtu=1500 mac-address=4C:5E:0C:B1:0B:78 arp=enabled 
      slaves=ether9,ether10,ether11,ether12
      mode=802.3ad primary=none link-monitoring=mii arp-interval=100ms arp-ip-targets="" mii-interval=100ms 
      down-delay=0ms up-delay=0ms lacp-rate=1sec transmit-hash-policy=layer-2 min-links=1 

 2  R name="bonding3" mtu=1500 mac-address=4C:5E:0C:B1:0B:78 arp=enabled 
      slaves=bonding1,bonding2 mode=active-backup primary=bonding2
      link-monitoring=mii arp-interval=100ms arp-ip-targets="" mii-interval=100ms down-delay=0ms up-delay=0ms
      lacp-rate=30secs transmit-hash-policy=layer-2 min-links=0

This works fine, until the router is rebooted. After a reboot, the two LAGs (bonding1, bonding2) are in “running” status, but “bonding3” doesn’t become active automatically.

The situation can be resolved by manually disabling and enabling the bonding interface.

Is there anything wrong with my configuration? Any other ideas?

Thanks in advance!

Oliver

2

You should probably use RSTP on the bridge itself, and let spanning tree do the active/passive failover behavior. I’m surprised that this even worked in the first place - are the two switches configured as a stack or are they standalone switches?
If stack, then just bond all 8 ports directly…
If standalone, then use spanning tree…

3

Thanks for the quick reply! The switches are standalone, I will see what happens if I just let RSTP do its work.

4

Just a short follow-up on this topic, in case somebody else has this problem. We ended up bridging the two LAGs and using RSTP to avoid loops. This works quite well.

Best regards

Oliver