Cisco AIR-CAP2702E-A-K9 dropping off network after 24 hour period

[derrick.plunk]

Hello!

I have recently inherited a network at at new job. The network consists of WLC 2500 (software version 7.6.120.0) with 5 AIR-CAP2702E-A-K9 AP (7.6.120.0) set in LAP mode and only 2 out of 5 APs work. What I have discovered is that if I power cycle the 3 AP points that are not working. They will work for about 24 hours before the LED flashing green and the APs points no longer showing up in the WLC GUI. I am not sure how to troubleshoot this issue. 

I have attached the screenshots of the log from CLI and WLC GUI. Any help is much appreciated!

[Leo Laohoo]

I am not sure how to troubleshoot this issue. 

Could be a faulty cable.

[derrick.plunk]

Thanks for the response, Leo. I checked that by taking the AP point down and plugging it directly into the main Cisco POE switch, and got the same issue. 

[Leo Laohoo]

Go to Management > Tech Support > AP Crash Logs. 

If there are items there that corresponds to the time and date the APs reboot themselves, then there's the answer. 

Another thing I might recommend is upgrade the firmware of the controller. 

[derrick.plunk]

No items in the AP crash logs. In fact, the AP Crash Logs screen was blank with no messages there. 

Do you have any tips on how to upgrade the firmware on the controller?

[derrick.plunk]

Here is a thought too. I am not sure the WAN controller is setup properly either. I have two connections going from the switch one is in a non-POE port and the other in a POE port. I am assuming this is for redundancy, but not sure.

Screenshot attached. 

[Scott Fella]

Depending on how your configuration is, the 4th port might not be required.  You can use LAG or not, but depends on your design of what you want to do. 

-Scott 

*** Please rate helpful posts *** 

[johnlloyd_13]

hi,

check if the access switch has enough power.

use the show power inline command to check.

try also replacing UTP cable and use the switch test cable tdr command if there's any open/short circuit.

[derrick.plunk]

Here is the log file from performing the cable test. I am curious about a few things here. 

Port 48 and Port 48 are connected to the APs that are currently working.

Port 45 and 46 are connected to APs that are not working. With the test results showing a fault at 0m meters. Does this mean the UTP cable going from switch to patch panel is bad or a bad termination on the back of the patch panel? 

Port 47 is connected to the third AP that is not working. I have this AP directly connected to the switch and used two different cables and got the same short error message. One of those cables being brand new. Maybe a faulty POE port on the AP?

[Scott Fella · Accepted solution]

We see this a lot in our environment also. To eliminate a bad switchport, connect another AP to that port or a laptop to that port (will not be able to verify PoE).  One thing also is to make sure there is no particles inside the AP Ethernet port and or the patch cable end.  A Ethernet cable tester can tell you exactly where the fault is, but if you don't have one, if you determine the AP and port is good, then replace the patch and re-terminate the rj45 end or re-punch the panel.

-Scott 

*** Please rate helpful posts *** 

[derrick.plunk]

Thanks for the tips, Scott! I used canned air to blow out the AP PoE port and switch port. The cable tdr test results came out as "Good". So I am letting the AP point sit for 24 hours to see if stays up now. 

Will report with results! 

[derrick.plunk]

Thanks for all the help with this guys! It looks like it was ports on the switch going bad. I moved the APs to different ports on the switch and they have been working great the past couple of days!

[Leo Laohoo]

TDR result shows the cable nearest the switch is faulty.