Wireless Access

Hi All,

One of our customers is getting a fairly large number of rx errors on the ports that their AP135's are patched into. They have a number of AP105's and AP135's but they're only seeing the rx errors on the switch ports that the AP135's are using.

The customer has a 3400 controller running ArubaOS 6.1.something. (I'll find out and update this).

I'm gonna check the ap tech-support details and the switch logs formore info but though it might be an interesting one...

There's a screengrab of the rx errors attached for your delightful viewing pleasure.

Cheers

James

AOS 6.1.3.7.

Interface Information
---------------------
Inter-| Receive | Transmit
----------------------------------------------------------------------------
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed reset
lo: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
eth0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
eth1:1625653911 15425530 0 0 0 0 0 0 312520437 1179754 0 0 0 0 0 0 0

------------

Ethernet Duplex/Speed Settings
------------------------------
Autoneg Speed (Mbps) Duplex Iface
------- ------------ ------ -----
on 10 Half eth0
on 100 Full eth1

------------------------------------

atapath Application Statistics
-------------------------------
NAPT Allocation Failures 0
NAT Allocation Failures 0
NAT TCP Sequence Failures 0
ICMP Errors Received 1848
ICMP Errors Denied 0
ICMP Errors Fragmented 0
ICMP Errors Generated 0

The customer has installed this AP and used eth1 to patch it into their 100Mbps PoE switch.

There are no details regarding the rx errors on the customers switch.

I've set one of their AP135's ethernet settings to 100Mbps full duplex to see if there are auto negotiate issues.

What would you guys recommend?

Use eth0 for starters?

Cheers
James

you should use eth0 really.  I've heard it said a few times that it makes no difference (if your wired-port config is the same), but my experience is different.

Those problems were mostly on the old AP70s, but following from that I always specify to the install guys to plug into eth0 now.

:smileyhappy:

This is just a thought, but I've seen this on my network as well. The fix for me was setting the MTU to 1500 in the AP System Profile. Some of my switches actually classify the issue for what it was, a jumbo frame, but others simply called it an input error. So that change may be worth a shot for you.

Thanks Guys. I've tried manually setting the MTU. Awaiting the results....

Manually setting the MTU did the trick. 

Thanks tbeachwcu. :smileyhappy:

You can see from the attached screenshot that the errors dropped away to nothing pretty much as soon as I made the change.

Glad that worked out for you!