1 gbps ---> 100 mbps randomly

[jadkins@skynetky.com]

I have a tower hop newly installed in my network. It is composed of two airfiber 5x units, netonix 12-250-dc switch, samlex dc power supply with battery backup and batterys, all new stuff. Tough cable runs to the hardware. Randomly one of the airfibers in this link will degrade port speed to 100 mbps. I can bounce the port and it stays back at 1 gbps for a few days then its back to 100mbps.

Any ideal what can cause this?

Here is a portion of the log:


Jan 10 04:23:45 STP: set port 2 to discarding
Jan 10 04:23:45 STP: set port 2 to learning
Jan 10 04:23:45 STP: set port 2 to forwarding
Jan 10 04:23:48 Port: link state changed to 'down' on port 2
Jan 10 04:23:48 STP: set port 2 to discarding
Jan 10 04:23:50 Port: link state changed to 'up' (1G) on port 2
Jan 10 04:23:51 STP: set port 2 to discarding
Jan 10 04:23:51 Port: link state changed to 'down' on port 2
Jan 10 04:23:52 Port: link state changed to 'up' (100M-F) on port 2
Jan 10 04:23:52 STP: set port 2 to discarding
Jan 10 04:23:52 STP: set port 2 to learning
Jan 10 04:23:52 STP: set port 2 to forwarding
Jan 10 04:37:21 Port: link state changed to 'down' on port 2
Jan 10 04:37:21 STP: set port 2 to discarding
Jan 10 04:37:28 Port: link state changed to 'up' (1G) on port 2
Jan 10 04:37:28 STP: set port 2 to discarding
Jan 10 04:37:28 switch[1413]: unexpected link change on port 2 100M-F
Jan 10 04:37:28 STP: set port 2 to learning
Jan 10 04:37:28 STP: set port 2 to forwarding
Jan 10 04:37:38 Port: link state changed to 'down' on port 2
Jan 10 04:37:38 STP: set port 2 to discarding
Jan 10 04:37:44 Port: link state changed to 'up' (1G) on port 2
Jan 10 04:37:44 STP: set port 2 to discarding
Jan 10 04:37:44 STP: set port 2 to learning
Jan 10 04:37:44 STP: set port 2 to forwarding
Jan 10 04:37:56 Port: link state changed to 'down' on port 2
Jan 10 04:37:56 STP: set port 2 to discarding

[sirhc]

In order of probability:
1) Bad Cable - replace cable
2) Bad End/Crimp - replace ends
3) Bad Radio - swap radio - If problem goes away bad radio or was bad connection
4) Bad port in switch - Move radio to another port if problem goes away then possible bad port or was lose connection
5) Interference around 100 Mhz - usually on brodcast towers

[jadkins@skynetky.com]

Their is an AM tower right next door to us at this site.

[sirhc]

Their is an AM tower right next door to us at this site.

Not sure if AM interferes with Ethernet cables communications like FM?

I would pursue the process of elimination I outlines above.

[slickdaddy0077]

I am having similar issues and they are random at a few different sites with Netonix WS-24-400B switches.

Issues:
Intermittent connectivity on random ports. Drops from 1000 down to 100 full (or) half duplex for some unknown reason. Sometimes bumping the port or re-booting the AP works but it's really a crap shoot that never really gives me any real idea where the trouble may be

One of the sites has an FM broadcast tower literally within 50 feet of some of the AP's in question that display these issues and the tower is broadcasting at 100,000 watts with the feed horns in direct horizontal wave of the AP's. We have taken great care to route the cables underground in conduit until they reach the structures (H frame & 1 wooden utility pole) No real 'tower at this location but we have approximately 30+ total radios at this site within close proximity of the FM transmitter which broadcasts at 97.5 Mhz.

The issues are random but I have noticed a trend. Radios mounted on the wooden utility pole are limited as to where to ground the radio / antenna. It seems that some some brands of radios are more picky than others as to how they perform through the Netonix switch and this may be due to ground loop at the equipment rack or a difference in ground potential from end to end. The cables are shielded and the drain wires are connected from end to end and cable testers do not detect any failts. They always pass as good but a slight variance in ground potential would not typically show up on a cable tester so I am suspicious and will need to use the Fluke DMM

I will be visiting our sites over the next week and will test them for ground loop and isolation, resistance potential and re-ground the switch chassis if necessary to see if this is in fact the cause or is contributing to the cause of ports dropping out randomly.

NOTE: The aggravating but typically effective fix for this has been to use a separate POE injector for the radio in question and power up the AP and then bypass the Netonix and go data directly into a Cisco catalyst switch as a temporary work-around fix.

Have you heard of ground loop and/or inconsistent ground potential at the switch chassis or equipment rack being an issue that could cause this but with random port isolation?

Thank you in advance and I will report my findings.

ML

[sirhc]

Never connect two device that have access to to 2 differnt grounding systems. Always bong towers ground to electrical service grounds as you will damage the equipment.

With that being said we have recently modified our designs to allow up to 40 amps of ground potential current to pass through the switch.

If you like you can RMA your switch and if you pay only the shipping we will add the modifications to your switch for free.

Here are some good posts on grounding:
viewtopic.php?f=30&t=1816
viewtopic.php?f=30&t=188
viewtopic.php?f=30&t=1429
viewtopic.php?f=17&t=1786&start=30#p13447
https://community.ubnt.com/t5/airFiber/ ... rue#M31070