VRRP transition logs

[Wendy2]

We have a site with 2 x MX in HA.  We are seeing  VRRP transition logs. The MX security appliances are frequently swapping between Master and Passive.

VRRP transition old_mode

detect prio 235

old_prio 235

elector_state master

last_state_change_reason master_down_timer

We are seeing MAC_flap logs on our Cisco switch.  No impact has been reported, but this is the only site we're seeing these errors on.  MX each have a lan port and wan port connected into separate switches in a stack.

Haven't been able to figure this out yet.

[alemabrahao]

How is your topology? This is usually due to an error in the topology design.

[Wendy2]

Thank you for your response.  MX_01 : Lan port into Cisco switch 1 trunk port.  WAN into internet router.  MX_02 : Lan port into Cisco switch 2 trunk port. This is a switch stack.

Packet capture is showing a loop between the VLAN Interface IP's

Checked spanning tree, and no errors.  This site is setup identically to another site with no MAC_flaps

[cmr]

Does the trunk port include VLAN1?

[Wendy2]

No, just the vlans setup on the MX

[cmr]

Is the Cisco switch stack the root bridge for all VLANs that are in the trunk?

Are there other switches in the LAN?

[Wendy2]

Yes, the core stack is the root bridge for all VLANS in the trunk.  We have two other access switches in the topology.  There is no impact to production as a result of this mac_flap so far.

[cmr]

As mentioned elsewhere, we did have the same for several firmware versions and a switch to DAC cables fixed it. Do you have spare SFP+ ports on the switches?

[alemabrahao]

Just to remind you that it is recommended that STP is enabled to avoid looping.

https://documentation.meraki.com/MX/Networks_and_Routing/MX_Layer_2_Functionality

[PhilipDAth]

>this is the only site we're seeing these errors on

Is this site running the same firmware versions as the other sites without issues?

[Wendy2]

Yes, it's identical to another site with the same topology.  MX_105 running the same versions 18.107.2

[cmr]

Are they MX250s and what firmware version are they running.  We had this issue with busy MX250s in concentrator mode where we used GLC-T modules in the WAN ports running at 1Gb.  Once we changed over to 10Gb DAC cables the transitions went away.

[Wendy2]

Thanks for the reply.  MX_105 running the same versions 18.107.2.  We're using ethernet ports

[cmr]

Are the Cisco switches running STP?

I'd upgrade to 18.107.10 at least and ideally a newer version.

How loaded are the MXs, more or less than the other site where there aren't errors?

[Wendy2]

The WAN ports lights are orange?  Is this correct?

[cmr]

Orange normally indicates that they are not running at the fastest port speed.  What does the dashboard report?

[Wendy2]

Dashboard says 1Gbps full duplex
Wondering if this issue is due to faulty physical cabling?

[cmr]

VRRP is over the LAN ports and the logs don't seem to show WAN down, unless that is also in there?

[Wendy2]

Site is fully up and functioning.  No issue reported about from the MAC_flaps I have noticed.  WAN is a separate thing, I agree