Route connection changes

[Neal]

Hi Meraki Guru!

I've been getting route connection changes on an MX84 that is tied to an MX100 over a Site to Site VPN Tunnel.

As soon as the route connection change hits, the tunnel goes down and I usually end up rebooting the MX84.

Once I do that, the tunnel comes back up. Since the ISP's equipment is tied into the MX84, would it be a valid

assumption that there is something going on with the connection to the ISP causing the route changes?

                                                                                                       Thanks......Neal Huhn

[MRCUR]

@Neal Here's another one @CarolineS! 

[MilesMeraki]

I hate to state the obvious but yeah it does seem related to your ISP. Are you seeing any other alerts in your MX's event log at that time that may be related?

[Neal]

No other messages. I will touch base with my ISP. 2nd question. Is there any other way to re-establish the tunnel without

rebooting devices at each end? (i.e. MX100 and MX84)....THANKS!

[Neal]

Hi MRCUR,

 I have 3 LAN connections going from my MX100 to a Cisco 6807-xl Core switch. My problem is I only have 1 of them up and active. Every time I bring up either of the 2 remaining with my 1 primary connection still up I start getting

'route connection change' events and the MX100 disconnects the routes. Note that neither the MX100 or the Core ports

go down. This is a brand new MX100 so it's not a hardware issue. Any thoughts on what's causing this and how I

can correct it?

                             Thanks!    Neal Huhn

[MRCUR]

The MX does not support creating a link aggregate on its LAN interfaces, nor does the MX participate in STP. So what you are doing is creating a loop between the MX and your Cisco core. Your Cisco core should be blocking one of the links with STP assuming you have STP configured on the Cisco core. 

[Neal]

So what you're telling me is I can only have 1 LAN connection between the MX100 and the Core?

[MRCUR]

I have only ever implemented one LAN link between an MX and a core switch sitting behind it. MX units support warm spare for redundancy. 

Perhaps @PhilipDAth has some insight on this. 

[Rudri]

We faced the similar issues, couple of spoke sites lost connectivity to servers hosted in AWS behind VMX HUB. When we verify the event logs - we observed below  event logs ,

Feb 19 17:21:41 BGP session established peer_ip: 10.184.54.1, local_as: 64512, remote_as: 64512
Feb 19 17:21:38 Route connection change peer_type: l3_vpn, peer: 0C:8D:DB:C0:17:80, connection_status: connected
Feb 19 17:21:33 VPN tunnel connectivity change vpn_type: site-to-site, peer_contact: 96.9.177.27:59038, connectivity: true
Feb 19 17:21:27 BGP session no longer established peer_ip: 10.184.54.1, local_as: 64512, remote_as: 64512 more »
Feb 19 17:21:27 BGP received notification peer_ip: 10.184.54.1, local_as: 64512, remote_as: 64512 more »
Feb 19 17:21:18 Route connection change peer_type: l3_vpn, peer: 0C:8D:DB:C0:17:80, connection_status: down

Can some one explain, is it because of this spoke sites been down ?