LACP goes down briefly when radio loses power?

[sbyrd]

I have an 8 port switch with active LACP to a Mk CCR-1009 router. Connected to the switch are also 4 AF5x radios configured as Master.

What is happening is when one of the Slave AF5x radios on another tower goes offline (hard reboot) I get the below alert from the switch and all devices on the switch stop pinging briefly.

LACP changed state to Active on port 8 (key 1) - from Elm St - Routed PoE Switch
LACP changed state to Active on port 7 (key 1) - from Elm St - Routed PoE Switch

Is this normal? How do I correct this?

[sirhc]

You're going to need to post up your Tabs for me to wrap my head around your installation.

Tabs I need:
Status
Ports
VLAN
LAG
STP
Device/Status
Device/Config

To upload pictures use the "upload attachment" link below the BLUE Submit button.

Please leave the Port Descriptions visible, I have no idea why people block that out, hopefully the port descriptions tell me what is on the ports.
Also upload a screen grab of Log showing the event with a description of how you caused it.

[sbyrd]

[sbyrd]

[sbyrd]

Here is my setup

Core router -> Elm Router (CCR-1009) -> LACP -> Netonix

On the Netonix in ports 3,4, and 5 I have UBNT AF-X radios programmed as Master. This switch is an aggregate switch for 3 backhauls that feed other towers in my network. My network is routed with the backhauls acting as bridges to connect the routers together.

Port 3 running fw 3.4b1
Port 4 running fw 4.0b1
Port 5 running fw 3.2

Before this switch was setup with just a single uplink to the Elm Router, but in an effort to distribute the pause frames between ports I setup this LACP a few months back.

What happened today that exposed this issue was the Slave radio to the Master AFX in port 4 (Slipher) went offline because the tower owner shutoff the main breaker to the tower. When it went offline I got a ping alert notification from The Dude that each device on this switch was Down and then a second later I got an alert it was Back online. Also at the same time I got the LACP alert from the netonix switch. The Mikrotik router's log does not show anything in regards the the LACP or ethernet ports going down.

As the power to the Slipher AFX Slave has been pulled 3 times today I have seen this happen each and everytime the Slave goes down.

[sbyrd]

The tower owner just cut the power to Slipher for the last time and as I was in the office saw something odd. Not only did I lose connection briefly to everything connected to the switch I also lost my Winbox connection to the Elm tower CCR-1009 router. However I never lost a Ping to the CCR-1009 router.

[sbyrd]

Also forgot I do not get an alert or lose communication when the Slave at Slipher comes back online.

[sirhc]

I am sorry Seth I still have not sat down and looked at this as I have been really busy. Will try and draw this out and look at it soon.

[sbyrd]

Let me know what I can do as it is now doing it at random times without any airfiber devices rebooting.

I cannot find what is triggering the random ones.

Any help would be appreciated?

[sbyrd]

I upgraded to 1.4.6 and still have this issue. The LACP goes down for a second for no noticeable reason. Is there anyone who can give me some help? I am getting desperate here?

Code: Select all

Jan 14 18:26:39 UI: Log cleared by admin
Jan 14 20:20:08 STP: set port 7 to discarding
Jan 14 20:20:08 STP: set port 8 to discarding
Jan 14 20:20:08 LACP: LACP changed state to Active on port 8 (key 1)
Jan 14 20:20:09 switch[1095]: LACP changed state to Active on port 8 (Uplink) (key 1)
Jan 14 20:20:09 STP: set port 11 to discarding
Jan 14 20:20:09 STP: set port 7 to discarding
Jan 14 20:20:09 STP: set port 8 to discarding
Jan 14 20:20:11 STP: set port 11 to learning
Jan 14 20:20:11 STP: set port 7 to learning
Jan 14 20:20:11 STP: set port 8 to learning
Jan 14 20:20:11 STP: set port 11 to forwarding
Jan 14 20:20:11 STP: set port 7 to forwarding
Jan 14 20:20:11 STP: set port 8 to forwarding