AP's lose connectivity same time every night

[johnny_5]

The past couple of nights our Cisco AP's 1242's go down approx at the same time. All of the AP's are connected to a 3560 POE with one more being connected to another 3560 POE. About 10-12 hours later they come back up and connectivity is restored.

It appears that the switches in question all blink green in unison really fast but every other device connected to the switches like PC's and IP phones are working as they should.

I'm thinking along the lines of a loop in the topology (3 POE switches, 1 non, 1 router).

I did a sh int cou err on the switch with the most AP's connected to it. Some of the interfaces show high "UnderSize" and some show 0 for the connected AP's so I don't know if that has anything to do with it.

The timing is the confusing thing - we have a second shift that runs from 3pm till 1am The problem starts around 7pm. My initial thoughts that one of the PC's or devices was causing the issue when they started their shift. The first shift starts at 5 am so they have to deal with the problem for 2-3 until they come back up.

Is there any way of isolating the possible issue?

[Leo Laohoo]

The past couple of nights our Cisco AP's 1242's go down approx at the same time. All of the AP's are connected to a 3560 POE with one more being connected to another 3560 POE. About 10-12 hours later they come back up and connectivity is restored.

Maybe you've got a time-based ACL or EnergyWise enabled? 

Are all the APs affected going to the same switch?

[johnny_5]

So 7 of the AP's are connected to one of the switches on the floor. The other AP is connected to another switch in the server room - switches are connected together via fiber.

There are no ACLs on the switch or EnergyWise. The problem has gotten progressively worse today with the AP's being down most of the day.

[Leo Laohoo]

So 7 of the AP's are connected to one of the switches on the floor. The other AP is connected to another switch in the server room - switches are connected together via fiber.

Let me get this straight, all the APs that are experiencing this problem are all connected to the same switch? 

if this is the case, what is the complete make/model of this switch?  

[johnny_5]

We have 10 AP's.

9 of them are connected to the switch A, with one more been connected to switch 2.

Switch A < - > Switch B are connected via fiber.

The one AP on switch B always goes down as well as some of the others on switch A.

Switch A : Cisco WS-C3560-48PS

Cisco IOS Software, C3560 Software (C3560-IPBASE-M), Version 12.2(35)SE5, RELEASE SOFTWARE (fc1)

Switch B: Cisco WS-C3560G-24PS

Cisco IOS Software, C3560 Software (C3560-IPBASE-M), Version 12.2(25)SEB2, RELEASE SOFTWARE (fc1)

[Leo Laohoo]

The one AP on switch B always goes down as well as some of the others on switch A.

Ok, so I'm going to concentrate on Switch B because Switch B is a 3560G and this specific model supports TDR.  The FastEthernet ports of the 3560 (Switch A) do not support TDR. 

Be aware that the IOS running on Switch B is extremely old and I don't even want to hazard a guess if TDR is supported on this IOS.  

Run the following commands: 

1.  Command:  test cable tdr interface <AP Port>;

2.  Wait for 61 seconds for the test to run

NOTE:  This is a 3560G and running a very old IOS.  The entire process will take 61 seconds (or more). 

3.  Command:  test cable tdr interface <AP Port>; and

4.  Post the complete output to #3.