Netonix Forums

I have a few WS-12-250Bs across my network, working great. They have all been working flawlessly with zero errors, but I recently installed a pair of AF5 links connecting me to my upstream provider, and one of them is showing errors.

Seeing as I am not sure what this Error column defines, can anyone point me in the right direction?
Does this mean something physically wrong with the cable? the radio? grounding? or the data coming over it?

At first the cable diagnostic was showing abnormal terminations, but we've changed out some ends and it shows "OK" now but the errors continue to rise. I reset the counter this morning, so 888 in the past 12 hours or so.

I am trying to diagnose some high pings and packet loss during our peak usage, 8pm-10pm. My upstream provider seems to think these port errors are the culprit. I am not so sure, but I have to fix this to eliminate that.

Thank you

The one data point I can provide is that I had one of the first batch 8 port units do the same thing with an even faster incrementing Rx drop counter than you are seeing and throughput decreased significantly when the problem started. It initially didn't behave this way. I replaced the switch with another unit and that issue virtually stopped. It still gets a few but I can watch it for a long time and never see one.

Can you please go to the status page & grab a screen shot so I can see what the system voltages are?

The cable test is a good tool but not perfect diagnostics.

The CRC and Rx Drop errors "usually" indicate a cable or bad jumper/end?

The oversize errors indicate something is running larger frames than the switch is configured to handle, you could increase the MTU on the switch but I would not use JUMBO frames on my backbone.

You are not by chance on a large tower with other antennas like AM/FM/TV or any other high power RF gear?

Jabber At times, Ethernet network interface cards and external transceivers also generate a problem called jabbering. This is when garbled bits of data are emitted within the frame sequence in a continuous transmission fashion. The packet length is usually more than 1,518 bytes.


Whats on port 4 - 15 watts is over the rating of that port?

With the current WS-12-250X model on the market you need to power any radios using more than .5A 24V or 48V on a port capable 48VH which provides of .75A at 24V and 48V.

This will be an "easy" diagnostic since port 1 and 2 both have the same equipment just swap ports and see if the errors follow the radio/cable but be sure to swap the jumpers and all if any.

ALWAYS TURN POE OFF before unplugging and plugging in then turn the POE back on.

I am concerned with the power draw on port 4 - Too HIGH for a standard 24V/48V port.

This is a cell tower that has 2 cell carriers about 15 feet above me.
Port 4 is a Rocket M5 with a standard ubnt 120deg sector, same as as ports 5 and 6. No customers on the sectors on 4 or 6 yet, and only 1 sub on port 5.

The errors continue to rise. I am also having some higher ping times at peak usage too, but no idea yet if those problems are related to this or not. I buy bandwidth from another wisp operator on this forum too, but I won't call him out yet

Are you using carrier grade shielded cable?
Have you tested you cables and connectors?

Here was my ping logs to akamai.net these evening. Not sure if these high ping times and packet loss is related to this or something else..

Using ubnt TC Carrier, we re did the ends on the wire twice, the wire diag shows ok now. but could always try that again.

Have you tried swapping the AF on ports 1 and 2 to determine if the problem is the port or the device?

I have not yet, I'll make a drive there maybe early in the morning and try that.