WISP Switch Will Not LAG 4 Ports

[krose@cottonwoodaz.gov]

I have two WISP switches in one building to power some Ubiquiti AirFiber and NanoBeam devices. One is WS-8-150-AC and the other is WS-12-250-AC. I am trying to connect both of these to a new Cisco SG550X-48P using four LAG ports each. I can connect one or two ports successfully, but as soon as I connect the third port it loses communication with all connected devices. I can no longer see any of the Ubiquiti devices and they lose internet connection. The WS-8-150-AC was working for a few days with four ports connected, then suddenly decided it would not longer work that way. The LAG ports are 1-4 on each switch. Any idea why this is happening?

[krose@cottonwoodaz.gov]

LAG settings on the 8-port switch:
Enable - Checked
Key - 1
Role - LACP-P
Timeout - Fast
Priority - 32768

Same settings on all four ports

[krose@cottonwoodaz.gov]

LAG settings on the Cisco SG550X-48P:
Each port set to Channel-Group 1 Mode Auto (Channel 2 for the 12-port switch)
Channel Group set to Trunk with a list of VLANs and a native VLAN

[Stephen]

I haven't seen this behavior myself, but there where a few issue's with LAG/LACP and STP that were solved sometime back, be sure to be on the latest firmware for the switch.

Digging around a bit, it appears that that Cisco switch might have some odd behavior with LAG/LACP as well, you may want to look here maybe you're seeing some sort of variation of this issue?

Also in case you're using SFP modules, many odd issue's seen on the forums have been traced to an SFP module that went bad.

[krose@cottonwoodaz.gov]

Thanks for your reply. I am not using an SFP module between the switches but it might be worth trying a single fiber connection instead of four Ethernet connections. I will check out the threads you mentioned. I sent this 12-port switch back for repair a while ago because it was not passing traffic at all right out of the box. I need to return this again for service.

On the 24th, the 12 port switch stopped passing traffic completely. I removed it from the network and went back to POE injectors instead.

[wtm]

We have the same problem with the Netonix switch and the Mikrotik routers. What we found was that you can NOT use the 802.ad type LAG with the Netonix switches. 2 LAG's will work, when you put on any additional ones, it all crashes. HOWEVER, if you select "Balanced rr" on the Mikrotik's it will work fine. I would assume that if you can change the type of LAG used on the CISCO routers, you will find the same thing. This problem started happening AFTER firmware version 1.5.1 (1.5.0 does not have this problem)

[krose@cottonwoodaz.gov]

Thanks for your reply. In the Cisco SG550X I have two options when adding a port to a channel group:

On - Add port without LACP

Auto - Add port with LACP

I will do some troubleshooting and see if the On setting makes any difference in setting up LAG.

Sadly the 12-port WISP switch was only using two ports in LAG when it crashed completely.

[krose@cottonwoodaz.gov]

Tried reprogramming the Cisco switch Port Channel mode to On instead of Auto. Just connected the LAG ports and it started downing several network devices. This was with no devices connected to the Netonix. I am sending this back and hoping for a refund.