WAP4410N stops responding with 2.0.0.5

[wheisler]

Just bought 2 of these (wish I had seen this thread before).  They shipped with the latest firmware already on them, so must be a very recent manufacture.  I am having the exact same problems as above.  It seems to be almost exactly 24 hours for me and the access point stops passing traffic.  At first the latency goes very high (2200ms - 3800ms) right before it stops passing traffic.  You can still ping the access point from the 802.11 side for a while, but nothing from the wired side, it just goes braindead.  Once I was able to get the management interface up over the wireless while it was in this state and rebooted from the management side.  It did not solve the issue, I had to unpower it and restart it (PoE Powered).  I have only configured one of them due to these issues.  I think they are going back.  Here is how the access point arrived (yes, I just rebooted again for the 8th time).

Hardware Version:	Rev.12
Software Version:	2.0.0.5
Local MAC Address:	00:13:60:FF:AF:F0
System Up Time:	0 hour:34 min:35 sec

[shepner2000]

As mentioned before, Ive been working with Cisco outside of this forum on the problem and its time for another update:

Last Ive heard, they are working on a true fix but for now I have a workaround for my specific environment.  Specifically, I have the WAP4410N access point and a SGE2000P Linksys switch (1.0.0.45 firmware).  The supplied recommendation was to turn off autogenotiation entirely.  Not leaving well enough alone, I found that if I set the switch to autonegotiate to *only* 1000F the problem still occurs but it isnt nearly as bad as normal.  When set to autonegotiate at *only* 100F I have not been able to cause the problem to occur.

Please remember that YMMV but on your switch, If your having the problem, try turning off auto negotiate and hardcode the port to 100F.  Unfortunately, I dont have any suggestions for unmanaged switches

hope this helps

stephen

[sreed]

Can I ask everyone Has anyone checked your power supply? I'm talking about the good old commercial power form your local power company or being delivered via POE.

I have one of these installed in my Home Office and had the 'hanging' problem very frequently till I confirmed we were having Power fluctuations/blinks. I switch mine to be directly connected to a small UPS about 3 months ago and haven't had one hang since.

It appears that these units are extremely sensitive to power fluctuations.

[orbital69]

Hi all, I thought that I would share my findings with you in regards to this issue. In my case it appears that the unit was overheating - since moving it away from sources of heat it has been operating very well as a WDS AP (not just a bridge). I still get the occasional drop out but the network resumes almost immediately whereas before once it dropped out it stayed that way until the unit was rebooted.

I have informed Cisco of my findings and also let them know that this isn't the first time I have come across a Linksys device that it particularly susceptible to heat.

Hope that this gives you something extra to play with in your troubleshooting!!

[unboundgil]

Stephen,

I appear to have the same problem that you have been having.  I set the AP's port on the switch to 10GF instead of auto-negotiate.  I thought this would fix the problem given the behavior I've been seeing in the unit and the solution I've been using up until this point.

BEHAVIOR

AP loses connection to LAN.  Clients can still connect to AP, but cannot communicate with the LAN.  Network clients connected to the LAN via ethernet on other parts of the network cannot connect to AP's management interface on LAN.  Switch indicates that the AP's LAN port is down.

SOLUTION (WORK AROUND)

AP does not need to be power cycled.  Unplug the LAN cable and then plug it back in.  AP is able to communicate with LAN again immediately.  This solution works consistently. 

THESIS

It appears that /something/ in the AP brings the switch's port down so that the switch thinks that nothing is connected.  Unplugging the AP's ethernet cable and plugging it back in breaks whatever loop the AP's software is trapped in.

I'm connected to a GS724T smart switch over cat5e.  Our old AP points did not have this problem.

I'd be happy to work with Cisco, too, if they need another Guinea Pig.  Hope this helps.

Gil