we have a WLC 4402 with 7 access points 1130 running in our office. in the access point statistics 802.11 MAC counters is see i high number of fragments and failures.
Tx Fragment Count 9209812
Tx Failed Count 279133
Multiple Retry Count 174745
ACK Failure Count 1446122
Multicast Rx Frame Count 234598
Tx Frame count 67297291
Multicast Tx Frame Count 137774
Retry Count 1540676
Frame duplicate count 59843
RTS failure count 94554
Rx Fragment count 0
FCS Error Count 22879679
WEP Undecryptable Count 384
any comment to the numbers ?
Wireless transmission is like a wired hub. One talks and the rest listens. It gets worst if your signal is weak because the client is either too far or too near (shadow).
Leo is on target with this answer. Are you having a specific issue or were you just looking at the stats?
This is normal. I know, your shaking you head right now and saying, HUH!?
Lets take RETRY'S for example. On a wired interface you dont see these that often UNLESS you have a wired issue. Perhaps a duplex miss match or some other issue. So typically speaking these counters are VERY low.
On wireless, way different story. Retry's are a way of life. Although, you shouldnt exceed a retry rate of 6%. Suppose for a minute your wireless client is on the edge of coverage. He sends a packet to the ap and the ap doesnt ACK back (remember when a STA sends a data frame an ACK is sent for acknowledgment). Since your on the edge of coverage you dont get that ACK. So what does the STA do. It resends marking the frame as a retry
If you arent having any network issues i woudlnt worry about it ...
If you want to test ... take an isolated test ap. Clear the counters and play with it for a while. Take a client on the edge and transfer an ISO and watch the counters.
on my Intel Client i see in the stats that i have about 8% packet loss. Signal strenght is -55dBM and Noise is -86dBM.
I think the packet loss is quite high.
Can you confirm my assumption?
I also forgot to mention the client saturation point or the number of clients talking to a particular AP. Too many clients talking to an AP can cause line errors too.
I'm working with a client that has a similar issue. They have a 4402 and about 30 APs. The FCS counters on all of the APs are growing at a fast pace (millions a day). I've opened a TAC case to try and determine if the errors I'm seeing are "normal" (within an acceptable range). The impression I have, thus far, is that it is not acceptable. Since the users are not happy with the performance of the applications which rely on the wireless network, we've decided to pursue this further. They had me dumb down the APs by making a few basic changes to the wlan and that did not reduce the rate of errors. Now they're focusing on transmit levels. The client only uses wireless from some thin clients and TAC wants me to find out what the max. transmit power level is for these thin clients. Wireless is not my cup of tea so I'm not going to try and get into a technical discussion. I'll only pass along what TAC is telling me (as I understand it). The APs can transmit at one of 8 power levels. Each power level is 1/2 the power level of the level above it. For example, PL1 would be 100 milliwatts, PL2 would be 50 milliwatts, PL3 is 25, etc. In any case, suppose that you have a device that has a maximum transmit power level of 30 milliwatts and your APs are set to transmit at 50 milliwatts. What will happen is that as the client devices get further and further out towards the edge of their transmission range, they will still be able to pick up signal from an AP but the AP will no longer be able to pick up signal from the client device. So, the client device will try to send data to the AP but the data will arrive at the AP with errors. In trying to explain this to me, they used voice as an example. They talked about how in wireless environments in which wireless IP phones exist, sometimes the phones will have a transmit level of about 30 milliwatts and what will happen is that users will experience 1-way audio issues because the stronger signal from the WAP will reach the phone but the phone signal will not reach the WAP. So as a phone user starts to roam further from the WAP, the user will still hear the caller on the other end but the caller on the other end will not hear the wireless phone user. As I understand it, the solution to that would be to reduce the transmit power on the APs to be more inline with the maximum transmit power of the lowest common denominator device on the network. The downside to that, though, would be more holes in coverage. You'd probably find yourself in a position where you'd have to deploy more APs to fill in the holes. At least that's the gist I'm getting from the discussion I had with TAC on Friday. In any case, if we end up doing somethnig that reduces the error rates I'll let you know what we end up doing.
I posted this on my site .. This may help with yout understanding... You are on target
I tracked my similar issue down to cabling issues.
SW03#show cable-diagnostics tdr interface Gig 1/0/5
TDR test last run on: August 17 17:12:39
Interface Speed Local pair Pair length Remote pair Pair status
--------- ----- ---------- ------------------ ----------- --------------------
Gi1/0/5 100M Pair A 123 +/- 4 meters Pair B Fail
Pair B 133 +/- 4 meters Pair A Fail
Pair C 152 +/- 4 meters Pair C Normal
Pair D 151 +/- 4 meters Pair D Normal
SW03#sho cdp neig Gig 1/0/5
Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge
S - Switch, H - Host, I - IGMP, r - Repeater, P - Phone
Device ID Local Intrfce Holdtme Capability Platform Port ID
Study Gig 1/0/5 123 AIR-LAP11 Fas 0
Please rate if this resolve your issue.
LOL, what is your issue? Do you have details?