Netonix Forums

Greetings, All.

So I've got a WS-12-400-AC on a tower. It's currently running 1.4.1, And is a Board Rev E. This is at our "Backhaul" level on the tower. It simply takes Fiber (SFP) in, And powers a couple of backhauls. Each backhaul is in it's own VLAN (Untagged on the Backhaul Port, Tagged down the fiber). And the switch's 'Management' vlan, As been changed to match our on-site management VLAN.

The backhauls include 1 Mimosa B5c, 1 Ligowave PTP-5-N PRO, 1 Radwin 2000c (I think that's what it is), And a single UBNT Rocket M5.

We're monitoring the switch via ICMP (Ping to confirm connectivity), As well as SNMP to pull stats like temperature, Voltage...etc. All of which experience quite a bit of packet loss. Even logged into the switch, I experience random pop ups of "Attempting to re-establish communication" or whatever it'll say in red at the top of the screen. I noticed this while the switch was running 1.3.x. I upgraded it to 1.4.1 and it still had the issue. As far as severity, It'll probably drop 10 out of 100 pings in the span of a minute or two. So this isn't like a single missed ping per-day or something small like that. Also, When it begins missing pings, It goes from zero latency, To 100% loss no high latency or anything preceding the loss. Nothing noteworthy in the switch logs. No ethernet errors are any port.

Now, I've worked with many a switch over the years. And this is by far not the first time I've seen control plane style loss on a switch. It's easy to confirm because all the connected devices have no loss / issues. This holds true with this particular netonix. Not a single issue to any device connected to the switch. However, This switch has no VLAN's..etc so it's a bit of a different scenario.

Being that this was our first heavily monitored Netonix switch, I brought up another one we were using as a simple switch at another location. This second location is a WS-12-250-AC, Firmware 1.4.2, Board Rev D. It has no such loss like the WS-12-400-AC.

So, My question is. Does anyone else see loss when pinging / monitoring their switch? This isn't exactly a "problem". But it's a bit annoying as you might imagine. Anyone else have any input?

Thanks!

There were some bugs found in v1.4.1 fixed in v1.4.2.

If you have UBNT discovery enabled or the Discovery Tab on v1.4.1 DISABLE them and see.

We are still working on eliminating some more bugs that can affect "some" people using Discovery protocols, the Discovery Tab, and SMTP alerts (if your mail server is slow to respond or mis-configured can cause issues).

So this is what I would do:
1) Upgrade to v1.4.2
2) If still having Issues try disabling Discovery Protocols, Discovery Tab, and possibly SMTP

Chris, Thanks for the quick reply.

Turned out to have nothing to do with the switch.

Found a device in the same VLAN gratuitously responding to ARP requests. Basically, The forward traffic would at times get switched down the wrong port of the router in front of the netonix switch, Which caused the loss that I saw.

Also explained why it had no loss to any of the connected equipment, As it was a different layer 2 (Different VLAN).