Wifi response fixed by reboot of AP or switch

Unanswered Question
Jul 25th, 2011

Hi,

For a remote location I manage the network, including wifi for production use. There are 7 switches (C2950), 11 accesspoints (C1240), 3 wifi-connected printers and 28 handheld wifi terminals (LXE) running a version of Windows CE. The terminals authenticate to an ACS using LEAP.

Periodically, the handheld terminals experience bad responsetimes within the application they use. This is a telnet application, very little data, connected to a SAP system where multiple other people make use of with a regular desktop SAP-client without response-problems.

The problems get fixed after a reload of either all switches or all accesspoints. This means if you reload all switches the problem gets fixed, but it also gets fixed if you reboot the accesspoints and leaving the switches alone.

The accesspoints are fairly new, about a year old. The switches are somewhat older, about 5 years. I recently upgraded the firmware on the switches, which did nothing to solve the issue. I can't find anything errorlike in the logs, before and after the reloads.

Where should I look for the problem?

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Average Rating: 0 (0 ratings)
jkeown Wed, 07/27/2011 - 06:22

Hi,

has the issue been observed in multiple physical locations, eg on multiple access points?

assuming that these ap's are in autonomous mode (I'm assuming this is the case since you didn't mention a controller), I'd start by capturing some snapshots of 'show dot11 assoc xxxx.xxxx.xxxx'.

Focus on one known problematic client, save the ouputs of the above command while things are healthy.

Once the issue happens again, take a few more snapshots of the same.

We're interested in the signal strength, error rates, etc.

here's an example:



1242-f#show dot11 assoc all

Address           : 0040.96b8.7a3f     Name             : JK-CUBE-DELL

IP Address        : 10.10.20.110       Interface        : Dot11Radio 0

Device            : CB21AG/PI21AG      Software Version : NONE

CCX Version       : 5

State             : Assoc              Parent           : self              

SSID              : wds-1242-f         VLAN             : 20

Hops to Infra     : 1                  Association Id   : 2

Clients Associated: 0                  Repeaters associated: 0

Tunnel Address    : 0.0.0.0

Key Mgmt type     : NONE               Encryption       : Off

Current Rate      : 54.0               Capability       : WMM ShortHdr ShortSlot

Supported Rates   : 1.0 2.0 5.5 6.0 9.0 11.0 12.0 18.0 24.0 36.0 48.0 54.0

Voice Rates       : disabled

Signal Strength   : -37  dBm           Connected for    : 665 seconds

Signal to Noise   : 50  dBm            Activity Timeout : 20 seconds

Power-save        : Off                Last Activity    : 0 seconds ago

Apsd DE AC(s)     : NONE

Packets Input     : 1319               Packets Output   : 12       

Bytes Input       : 77378              Bytes Output     : 1142     

Duplicates Rcvd   : 18                 Data Retries     : 2        

Decrypt Failed    : 0                  RTS Retries      : 0        

MIC Failed        : 0                  MIC Missing      : 0        

Packets Redirected: 0                  Redirect Filtered: 0    



You may also want to run a carrier busy test while the problem is evident:

IOS-AP#dot11 dot11Radio 0 carrier busy

*Apr  5 12:25:25.933: %LINK-5-CHANGED: Interface Dot11Radio0, changed state to reset

*Apr  5 12:25:26.933: %LINEPROTO-5-UPDOWN: Line protocol on Interface Dot11Radio0, changed state to down

Frequency  Carrier Busy %

---------  --------------

   2412         25

   2417         35

   2422         43

   2427         35

   2432         39

   2437         51

   2442         46

   2447         42

   2452         25

   2457         29

   2462         66

*Apr  5 12:25:31.461: %LINK-3-UPDOWN: Interface Dot11Radio0, changed state to up

*Apr  5 12:25:32.461: %LINEPROTO-5-UPDOWN: Line protocol on Interface Dot11Radio0, changed state to up

IOS-AP#show controller dot11Radio 0

!

interface Dot11Radio0

Radio AIR-AP1131G, Base Address 001c.f905.9280, BBlock version 0.00, Software version 6.30.13

Serial number: GAM11280R2G

Number of supported simultaneous BSSID on Dot11Radio0: 16

Carrier Set: Americas (US) (-A)

Uniform Spreading Required: No

Configured Frequency: 2437 MHz  Channel 6

Active  Frequency: 2447 MHz  Channel 8 20MHz

Allowed Frequencies: 2412(1) 2417(2) 2422(3) 2427(4) 2432(5) 2437(6) 2442(7) 2447(8) 2452(9) 2457(10) 2462(11)

Listen Frequencies: 2412(1) 2417(2) 2422(3) 2427(4) 2432(5) 2437(6) 2442(7) 2447(8) 2452(9) 2457(10) 2462(11) 2467(12) 2472(13) 2484(14)

Isegrimm24 Tue, 08/02/2011 - 23:45

It's been observed in multiple locations. The AP's are indeed in autonomous mode.

I've checked the statistics and during periods where no problems are reported I saw one terminal with received duplicates (about 7% of the total packet count) and retries. Other terminals showed much less or no errors at all.

The average signal-to-noise ratio of the terminals I observed was around 40 dB.

Signal strength varied from -42 dBm to -65dBm.

These figures show no real problems in my opinion. Any thoughts?

Leo Laohoo Tue, 08/02/2011 - 23:52

If the WiFi is fixed only by rebooting the AP, maybe you need to upgrade the IOS?

What IOS version are you running on your AP and your switch?

handheld terminals experience bad responsetimes within the application they use

Only the handheld terminals?  How about laptops?  Sounds like your application could be acting up.  Have you tried to "wireshark" the data exchange between the application and the handhelds?

Isegrimm24 Wed, 08/03/2011 - 00:18

I had the same thought, so I upgraded the IOS on the switches to the latest available but this had no effect.

The AP's are running on 12.4(10b)JDA3, it's from 2009 but still listed as the latest GD release.

I don't think the application could be acting up. Other users (wired desktops) experience no problem using the same SAP system. Also, I wouldn't expect such a problem could be fixed by rebooting the switches or AP's.

Leo Laohoo Thu, 08/04/2011 - 18:33
The AP's are running on 12.4(10b)JDA3, it's from 2009 but still listed as the latest GD release.

That's an OLD version IOS.  Latest one is 12.4.(25d)JA.

Isegrimm24 Sun, 08/28/2011 - 05:52

Yes, I know it is a rather old firmware, but since it is still the latest GD release I initially thought it best to leave it in place. However, to rule out the firmware as a cause I have upgraded it about two weeks ago to the last one available for this model, but the problems seem to persist. We also had an external company do a fieldtest to help find the problem, but they could not find any problem with the wireless network.

Sent from Cisco Technical Support iPad App

Isegrimm24 Tue, 08/02/2011 - 23:57

And this is a typical carrier busy output for the AP's. I could only check the ones without assocations ofcourse, but if this is an indication for the usage on the remaining AP's there don't seem to be any apparent cause so far.

Frequency  Carrier Busy %

---------  --------------

   2412          2

   2417          2

   2422          2

   2427          0

   2432          3

   2437          1

   2442          0

   2447          0

   2452          0

   2457          0

   2462          0

   2467          0

   2472          0

Actions

Login or Register to take actions

This Discussion

Posted July 25, 2011 at 1:56 AM
Stats:
Replies:7 Avg. Rating:
Views:767 Votes:0
Shares:0
Tags: No tags.

Discussions Leaderboard