Does the ether connection status change when a cellular change happens ? (MG51)

[thomasthomsen]

Im trying to figure out whats going on here in this eventlog (see below).

The MG actually have to sims from different providers.

Auto failover is enabled.

SIM1 (as you can see) is TeliaDK

SIM2 is TDC

MG is getting power from PSU, not PoE

In Denmark we have a very good idea of what ISP is where (mastedatabasen.dk will tell you)

When we look at this customers site, they are sitting, "more or less", in the middle of 3 different TeliaDK cells.

Only 1 TDC cell is nearby

This site looses connection throughout the day, and it matches pretty well with the below log.

So I was wondering, does the below log show roaming between these cells ?

Because it never fails over to the secondary SIM.

But why does the ethernet connection go up and down ?

This is very strange behaviour I think.

Oct 25 14:22:28			Cellular	Cellular connection up	local: 100.83.124.240, remote: 100.83.124.241, provider: TeliaDk, connection: AUTO
Oct 25 14:18:46			Status	Ethernet port carrier change	device: port1, carrier: true
Oct 25 14:18:29			Status	Ethernet port carrier change	device: port1, carrier: false
Oct 25 14:18:27			Appliance status	Primary uplink status change	is_cellular: 2
Oct 25 14:17:51			Status	Ethernet port carrier change	device: port1, carrier: true
Oct 25 12:05:18			Cellular	Cellular connection up	local: 100.83.162.83, remote: 100.83.162.84, provider: TeliaDk, connection: AUTO
Oct 25 12:02:37			Status	Ethernet port carrier change	device: port1, carrier: true
Oct 25 12:02:20			Status	Ethernet port carrier change	device: port1, carrier: false
Oct 25 12:02:18			Appliance status	Primary uplink status change	is_cellular: 2
Oct 25 12:01:43			Status	Ethernet port carrier change	device: port1, carrier: true
Oct 25 09:38:17			Cellular	Cellular connection up	local: 100.83.217.186, remote: 100.83.217.185, provider: TeliaDk, connection: AUTO
Oct 25 09:35:41			Status	Ethernet port carrier change	device: port1, carrier: true
Oct 25 09:35:19			Status	Ethernet port carrier change	device: port1, carrier: false
Oct 25 09:35:17			Appliance status	Primary uplink status change	is_cellular: 2
Oct 25 09:34:41			Status	Ethernet port carrier change	device: port1, carrier: true
Oct 25 07:37:24			Cellular	Cellular connection up	local: 100.67.153.179, remote: 100.67.153.180, provider: TeliaDk, connection: AUTO
Oct 25 07:34:48			Status	Ethernet port carrier change	device: port1, carrier: true
Oct 25 07:34:26			Status	Ethernet port carrier change	device: port1, carrier: false
Oct 25 07:34:24			Appliance status	Primary uplink status change	is_cellular: 2
Oct 25 07:33:49			Status	Ethernet port carrier change	device: port1, carrier: true

[thomasthomsen]

This is the MX log for ethernet port carrier change:

Oct 25 14:19:06		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 14:19:02		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 14:17:49		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 14:17:23		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 14:08:25		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 14:08:20		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 14:03:37		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 14:03:34		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 12:02:25		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 12:02:22		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 12:01:03		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 12:00:28		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 11:57:32		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 11:57:30		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 09:35:24		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 09:35:21		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 09:34:08		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 09:34:03		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 09:33:59		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 09:33:42		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 09:32:21		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 09:32:18		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 07:34:31		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 07:34:27		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 07:33:16		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 07:32:22		Status	Ethernet port carrier change	device: port0, carrier: false
Oct 25 07:30:35		Status	Ethernet port carrier change	device: port0, carrier: true
Oct 25 07:30:32		Status	Ethernet port carrier change	device: port0, carrier: false

[alemabrahao]

Generally Ethernet port carrier change indicates a physical problem, have you checked whether all connections are ok?

[thomasthomsen]

I would completely agree.

But Im just wondering why they seem to match up with the : "Primary uplink status change" of the MG also.

This seems strange to me. Why would the cellular interface change when the ethernet link goes down on an MG (not PoE powered of course).

[javalins]

It shouldn't, what code version are you running?

[thomasthomsen]

MG 3.100

MX 17.10.9 - MX68

[PhilipDAth]

This sounds like a support case job ...

This does not look like correct behaviour to me.  It seems like a software bug (to me).

[thomasthomsen]

Yeah. I might open a case.

I just want to be sure everything is connected correctly first (also a damm shame that there is no interface counters on the MX, its on my wishlist 🙂 )

[javalins]

Please do, if this is a defect it will need to be routed to engineering. 

[javalins]

Did a support ticket ever get opened on this?

[harmankardon]

@thomasthomsen Did you ever come to a resolution on this?

[thomasthomsen]

Sorry for the late reply. No - we kinda abandoned the troubleshooting on this point, because the prime problem was low signal strength for the current posistion of the MG.

I do not have one in my lab, so I had no way of doing additional testing.