Bonding Question

RouterOS General
1

Morning! I am having some problems with my bond interface, the settings are below.

I have had a active backup bond for about 6 months now with no issues, and now I am starting to see the switch start throwing errors.

<190> Mar 17 10:10:39 L0 Distro1 - ProdSW_D001-1 FDB[dtlAddrTask]: fdb.c(690) 1232709 %% INFO MAC_MOVE: Mac 78:9A:18:2B:BA:1A in VLAN: 1 is overwritten from entryType 1 to 1 and port 1/0/35 to 1/0/40
<190> Mar 17 10:10:38 L0 Distro1 - ProdSW_D001-1 FDB[dtlAddrTask]: fdb.c(690) 1232708 %% INFO MAC_MOVE: Mac 78:9A:18:2B:BA:1A in VLAN: 1 is overwritten from entryType 1 to 1 and port 1/0/40 to 1/0/35
<190> Mar 17 10:10:38 L0 Distro1 - ProdSW_D001-1 FDB[dtlAddrTask]: fdb.c(690) 1232707 %% INFO MAC_MOVE: Mac 78:9A:18:2B:BA:1A in VLAN: 1 is overwritten from entryType 1 to 1 and port 1/0/35 to 1/0/40
<190> Mar 17 10:10:37 L0 Distro1 - ProdSW_D001-1 FDB[dtlAddrTask]: fdb.c(690) 1232706 %% INFO MAC_MOVE: Mac 78:9A:18:2B:BA:1A in VLAN: 1 is overwritten from entryType 1 to 1 and port 1/0/40 to 1/0/35

One link of the bond is on my main core switch, and the second link is on my secondary core switch. (STP 0 and STP 4096). The ARP IPs are the 2 core switches. I can’t figure out why all of a sudden the switches keep fussing about the MAC addresses moving around. The router doesn’t say i’ts switching interfaces back and forth.

For now, I have withed it to MII/Link Monitoring and the switch messages have stopped. But in my mind the security of ARP to know data can pass through a link is better than just a connection up check.

[admin@MikroTik] > /interface bonding print
Flags: X - disabled; R - running 
 0  R ;;; Main Bond
      name="bonding1" mtu=1500 mac-address=78:9A:18:2B:BA:1A arp=enabled arp-timeout=auto slaves=sfp-sfpplus2,sfp-sfpplus3 mode=active-backup 
      primary=sfp-sfpplus2 link-monitoring=arp arp-interval=1s arp-ip-targets=10.201.1.5,10.201.1.6 mii-interval=100ms down-delay=0ms up-delay=2s 
      lacp-rate=30secs transmit-hash-policy=layer-2-and-3 min-links=0




[admin@MikroTik] > /interface bonding print
Flags: X - disabled; R - running 
 0  R ;;; Main Bond
      name="bonding1" mtu=1500 mac-address=78:9A:18:2B:BA:1A arp=enabled arp-timeout=auto slaves=sfp-sfpplus2,sfp-sfpplus3 mode=active-backup 
      primary=sfp-sfpplus2 link-monitoring=mii arp-interval=1s arp-ip-targets=10.201.1.5,10.201.1.6 mii-interval=100ms down-delay=0ms up-delay=2s 
      lacp-rate=30secs transmit-hash-policy=layer-2-and-3 min-links=0
2

I know it’s not the answer to your question, but for what reason are you not using an active protocol (802.3ad)?

3

It may be inexperience on my part. I was under the impression that both/all links on a LAG have to be connected to the same hardware. This is redundancy between my two core switches, so one link from router would be on each switch.


–Dan

4

You are right. But if both switches are MikroTik CRS3xx, you could do MLAG + 802.3ad, which also might solve the problem.

5

The bond XR option should only be used with some smart (but dumb) managed switches that only have LAG option. Concur with patrick, 802.ad is the way to go.